Datasets:

lambdaofgod

/

pwc_repositories_with_dependencies

like 0

0bserver07/One-Hundred-Layers-Tiramisu

['semantic segmentation']

['The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation']

camvid_data_loader.py helper.py fc-densenet-model.py model-dynamic.py model-tiramasu-103.py model-tiramasu-67-func-api.py model-tiramasu-56.py train-tiramisu.py model-tiramasu-67.py load_data Tiramisu normalized one_hot_it Tiramisu Tiramisu Tiramisu Tiramisu Tiramisu step_decay rollaxis print len normalized append range one_hot_it zeros float32 equalizeHist zeros range pow floor

0bserver07/One-Hundred-Layers-Tiramisu

101vinayak/Neural-Style-Transfer

['style transfer']

['A Neural Algorithm of Artistic Style']

images2gif.py checkImages writeGif get_cKDTree readGif NeuQuant GifWriter intToBin append uint8 astype copy int int checkImages hasattr handleSubRectangles GifWriter writeGifToFile convertImagesToPIL open fromarray asarray seek tell convert append enumerate open

# Neural-Style-Transfer Neural style transfer is an optimization technique used to take two images—a content image and a style reference image (such as an artwork by a famous painter)—and blend them together so the output image looks like the content image, but “painted” in the style of the style reference image. Implementation of https://arxiv.org/abs/1508.06576 Research paper from Cornell Univ. STYLE TRANSFER PROCESS: <h3 align="center"> <img src="result.gif"> </h3> The process follows the use of VGG19 model of CNN which is used for content and style extraction features in the network. <img src="VGG19.png">

12kleingordon34/NLP_masters_project

['word embeddings']

['Gender Bias in Contextualized Word Embeddings']

process_winogender_data.py main process_wino_data process_occ_stats float int list items replace append process_wino_data process_occ_stats

# NLP_masters_project Code base used for NLP project 2020. By Daniel de Vassimon Manela, Boris van Breugel, Tom Fisher, David Errington --- ## Contents * `process_ontonotes.ipynb`: Loads the Ontonotes Release 5.0 data from [Github](https://github.com/yuchenlin/OntoNotes-5.0-NER-BIO.git), and processes raw data into a suitable formatting for modelling. Notebook is also dependent on data loaded from [UCLA NLP Group](https://github.com/uclanlp/gn_glove.git). Both these files are loaded within the notebook. Running notebook back-to-front returns `original_data.csv` and `flipped_data.csv`, which are used in training a BERT Masked Language Model. * `BERT_fine_tuning_full_ontonotes.ipynb`: Loads the output of `process_ontonotes.ipynb` and trains a BERT Masked Language Model using [Huggingface](https://github.com/huggingface/transformers) directory as described in Section 4.3 of our paper. Be sure to correctly comment/uncomment out the lines for loading the data. Data Augmented models require loading in both `original_data.csv` and `flipped_data.csv`. To train regular unaugmented finetuned models, only load `original_data.csv`. * `bias_analysis.ipynb`: Analyses streo and skew bias in out-of-the-box ELMo, BERT, distilBERT and RoBERTa using [WinoBias dataset](https://arxiv.org/pdf/1904.03310.pdf). Also includes Online Skewness Mitigation for different BERTs, visualisations of professional embedding bias, and compares bias in BERT to industry statistics. ## Extra Dependencies

131250208/TPlinker-joint-extraction

['relation extraction']

['TPLinker: Single-stage Joint Extraction of Entities and Relations Through Token Pair Linking']

tplinker_plus/config.py common/utils.py tplinker/config.py tplinker/train.py tplinker_plus/train.py setup.py common/components.py tplinker/tplinker.py preprocess/__init__.py tplinker_plus/tplinker_plus.py LayerNorm HandshakingKernel DefaultLogger Preprocessor DataMaker4BiLSTM TPLinkerBert HandshakingTaggingScheme DataMaker4Bert TPLinkerBiLSTM MetricsCalculator train_step train_n_valid MyDataset bias_loss valid_step DataMaker4BiLSTM TPLinkerPlusBert TPLinkerPlusBiLSTM HandshakingTaggingScheme DataMaker4Bert MetricsCalculator train_step sample_equal_to train_n_valid MyDataset valid_step to CrossEntropyLoss get_sample_accuracy backward rel_extractor zero_grad loss_func step get_rel_cpg get_sample_accuracy join format valid state_dict glob print save train range len add format set_trace set repeat gather long get_cpg long

# TPLinker **TPLinker: Single-stage Joint Extraction of Entities and Relations Through Token Pair Linking** This repository contains all the code of the official implementation for the paper: **[TPLinker: Single-stage Joint Extraction of Entities and Relations Through Token Pair Linking](https://www.aclweb.org/anthology/2020.coling-main.138.pdf).** The paper has been accepted to appear at **COLING 2020**. \[[slides](https://drive.google.com/file/d/1UAIVkuUgs122k02Ijln-AtaX2mHz70N-/view?usp=sharing)\] \[[poster](https://drive.google.com/file/d/1iwFfXZDjwEz1kBK8z1To_YBWhfyswzYU/view?usp=sharing)\] TPLinker is a joint extraction model resolved the issues of **relation overlapping** and **nested entities**, immune to the influence of **exposure bias**, and achieves SOTA performance on NYT (TPLinker: **91.9**, TPlinkerPlus: **92.6 (+3.0)**) and WebNLG (TPLinker: **91.9**, TPlinkerPlus: **92.3 (+0.5)**). Note that the details of TPLinkerPlus will be published in the extended paper, which is still in progress. **Note: Please refer to Q&A and closed issues to find your question before proposed a new issue.** - [Model](#model) - [Results](#results) - [Usage](#usage) * [Prerequisites](#prerequisites) * [Data](#data)

15saurabh16/Multipoles

['time series']

['Mining Novel Multivariate Relationships in Time Series Data Using Correlation Networks']

COMET_fMRI_Multiple_Parameters.py COMET_Climate.py bronk_kerb.py empirical_observations.py SignifTesterMultipole.py quick-cliques/utils/edges2snap.py quick-cliques/utils/invertdimacs.py Misc_Modules.py get_graph_in_txt.py COMET_ADVANCED.py RandomSearch_Climate.py COMET_fMRI.py quick-cliques/utils/edge2dimacs.py quick-cliques/utils/edges2graph.py BruteForce_fMRI.py BruteForce_Climate_New.py COMET.py compare_with_brute_force.py quick-cliques/utils/dimacs2edges.py completeness_eval.py generate_random_ts.py RandomSearch_fMRI.py Comet_scalability_analysis.py BruteForce_fMRI_New.py GraphLASSObaseline.py BruteForce_Climate.py convert_txt_to_mat_lasso.py COMET_Climate_Multiple_Parameters.py CLIQUE_multipoles_algorithm.py compare_multipole_algorithms.py COMET_Synthetic_Multiple_Parameters.py levg_null_dist_expt3.py degeneracy_ordering find_cliques_pivot find_cliques get_inputs_for_groups brute_search mkdirnotex find_max_edgewt brute_search_parallel2 brute_search_group get_maxedgewt_vec get_inputs_for_groups brute_search mkdirnotex find_max_edgewt brute_search_parallel2 brute_search_group get_maxedgewt_vec mkdirnotex mkdirnotex remove_neg_cliques extract_multipole find_good_multipoles remove_duplicate_cliqs_2 remove_duplicate_cliqs construct_GraphF remove_g2_cliques get_supersets CLIQ_ALGO get_inputs_for_groups extract_multipoles_from_cliq cliq_brute_search COMET_EXT find_good_multipoles_complete_parallel extract_multipoles_from_cliqgroup get_inputs_for_groups extract_multipoles_from_cliq cliq_brute_search COMET_EXT extract_multipoles_from_cliqgroup find_good_multipoles_complete_parallel remove_non_maximals mkdirnotex mkdirnotex mkdirnotex mkdirnotex mkdirnotex mkdirnotex get_sorted_mps find_MP_in_MPList get_MPs_of_M1_missed_by_M2_NOLEV get_MPs_of_M1_missed_by_M2 get_MPs_of_M1_missed_by_M2_ALTER gen_dist_kxk_mats_edge_range get_lev_and_levg get_lev_minors gen_dist_kxk_mats_min_edge mkdirnotex make_sym_matrix generate_std_rand_ts_corrmat get_graph_in_txt filter_multipoles get_lev_and_levgs_all_cands get_multipole_cands pval_anal3_SLP get_null_dist_expt3 pval_anal3_fMRI get_inputs_for_groups get_lev_levg_and_weakest_pole key_generator remove_redundant_multipoles_alter_parallel_recursive get_lev_and_levg find_next_level_multipoles remove_redundant_multipoles_alter get_lev_minors remove_redundant_multipoles_alter_group find_multipoles_for_P2 remove_redundant_multipoles_alter_parallel object_to_list_arr remove_redundant_multipoles get_inputs_for_groups brute_search random_search_parallel2 generate_final_output mkdirnotex brute_search_group get_inputs_for_groups brute_search random_search_parallel2 generate_final_output mkdirnotex brute_search_group pval_anal_fMRI get_pval_multipole get_lev_and_levg_all_wins get_lev_and_levg_group pval_parallel pval_anal_SLP get_pval_multipole_group get_multiple_groups pval_anal_model_SLP degeneracy_ordering remove find_cliques_pivot union set add intersection keys remove add difference iter intersection append next union pop remove defaultdict set add reverse append len dirname makedirs size arange len str list time combinations get_lev_and_levg print append range list time get_lev_and_levg zip append max Pool str list map append sum next range factorial get_inputs_for_groups close zip combinations print min divide repeat len triu_indices argmin eig where max zeros len range find_max_edgewt append issubset range len tolist array range list tolist append array range len list tolist append array range len int list min tolist copy split append array range len list min tolist copy append array range len list get_lev_and_levg min argmin divide mean eye abs range extract_multipole str list time print zeros range len remove_neg_cliques str time find_good_multipoles print tolist remove_redundant_multipoles_alter copy find_cliques construct_GraphF range combinations list get_lev_and_levg min tolist divide mean eye append abs array range len int cliq_brute_search min range len extract_multipoles_from_cliq time list str get_lev_and_levg print zip range len get_inputs_for_groups list print close map repeat zip append Pool range len str time print tolist copy find_cliques find_good_multipoles_complete_parallel remove_duplicate_cliqs remove_redundant_multipoles_alter_parallel construct_GraphF range len extract_multipoles_from_cliq list sorted sort append range len format get_graph_in_txt chdir getcwd system remove_non_maximals mkdir remove_duplicate_cliqs_2 tolist argsort append array range len get_sorted_mps list set_trace set append range len get_sorted_mps list issubset set append range len list print issubset set append range len len issubset range set zeros triu_indices pop list eig min zeros range eig get_lev_minors min triu_indices get_lev_and_levg argmin eig rand where append max range make_sym_matrix gen_dist_kxk_mats_edge_range str arange print size range normal int zscore matmul cholesky zeros range format len write close range open list min append abs range len range get_lev_and_levg len append range len list get_lev_and_levg corrcoef sample zeros range len str time format print getcwd get_null_dist_expt3 append loadmat range time format print getcwd transpose size get_null_dist_expt3 append loadmat range list sort issubset zip append zeros range len list remove_redundant_multipoles_alter zip get_inputs_for_groups list min close map remove_redundant_multipoles_alter zip Pool len get_inputs_for_groups list format print min close map remove_redundant_multipoles_alter zip Pool len str list time print sort zip append zeros array range len argmin eig get_lev_minors min reverse list get_lev_and_levg sort set append range len list format print close map set repeat zip append sum Pool range str list time format generate_final_output print min close map choice repeat zip append sum Pool range len str format print getcwd savemat mkdirnotex append remove_non_maximals range len print size arange len zeros range get_pval_multipole len pop list get_lev_levg_and_weakest_pole get_lev_and_levg size corrcoef sample zeros float range len list get_lev_and_levg corrcoef zip zeros range len max list close map repeat array zip get_multiple_groups zeros Pool range len time list format concatenate print tuple close delete map repeat zip get_multiple_groups zeros Pool max range len str get_lev_and_levg_all_wins pval_parallel savemat append expanduser loadmat range str getcwd get_lev_and_levg_all_wins pval_parallel append loadmat range getcwd transpose size get_lev_and_levg_all_wins pval_parallel append loadmat range

# Multipoles This repository contains the code that we used to find "multipoles", a new class of multivariate relationship patterns in time series data in datasets from climate and neuroscience domain. See [1],[2] for further details. <b>Instructions to run the code:</b> 1) The code to obtain all the empirical observations discussed in section 4.1 of techincal report can be obtained using empirical_observations.py. 2) Please keep all the codes and data files in the same directory. 3) There are two implementations of COMET algorithm; COMET.py and COMET_ADVANCED.py. The two implementations mainly differ in the choice of algorithm used to solve the clique enumeration problem in one of the intermediate steps. Script COMET.py uses bronk_kerb.py, a python implementation of bronk kerbosch's algorithm (1973), whereas COMET_ADVANCED.py uses the C++ implementation (in quick-cliques folder) provided by authors (available at https://github.com/darrenstrash/quick-cliques). contains the primary module COMET_EXT which can be called to find all multipoles in a given dataset. COMET_EXT takes five inputs:<br> CorrMat: Correlation matrix of the entire dataset. <br> sigma: minimum threshold on linear dependence of desired multipoles. <br> delta: minimum threshold on linear gain of desired multipoles.<br> edge_filt: same as parameter \mu of CoMEtExtended (see <a href = "https://www.researchgate.net/publication/323129038_Mining_Novel_Multivariate_Relationships_in_Time_Series_Data_Applications_to_Climate_and_Neuroscience"> technical report </a> for further details). Higher values of edge_filt will recover more multipoles but will increase computational time and vice-versa.<br>

1980x/ABAW2020DMACS

['facial expression recognition']

['Affect Expression Behaviour Analysis in the Wild using Spatio-Channel Attention and Complementary Context Information']

dataset/affectwild2_dataset.py models/losses.py test_affwild2.py train_affwild2_expw_affectnet.py util.py dataset/sampler.py train_affwild2.py dataset/affectwild2_expw_affectnet.py models/attentionnet.py models/resnet.py validate statistic AverageMeter accuracy save_checkpoint adjust_learning_rate main train val_accuracy load_state_dict _read_path_label PIL_loader change_emotion_label_same_as_affectnet default_reader ImageList switch_expression _read_path_label default_reader_expw PIL_loader change_emotion_label_same_as_affectnet default_reader_affectnet change_emotion_label_of_expw_same_as_affectnet default_reader ImageList switch_expression ImbalancedDatasetSampler RegionBranch eca_layer Attentiomasknet count_parameters extract_patches_attentivefeatures SharedAttentionBranch AttentionBranch main FocalLoss LDAMLoss focal_loss norm_angle count_parameters ResNet resnet50 Bottleneck resnet152 sigmoid conv3x3 resnet34 resnet18 main BasicBlock resnet101 validate batch_size count_parameters warn SGD pretrained DataLoader adjust_learning_rate save_checkpoint dataset max str ImbalancedDatasetSampler load_state_dict AttentionBranch parse_args to ImageList sum range format TestList resnet50 Compose test start_epoch add_param_group resume item power get_cls_num_list load RegionBranch print imagesize predict_test isfile train epochs array len zero_grad region_model to range cat update format size mean item basemodel enumerate time attention_model criterion backward print AverageMeter accuracy step len f1_score recall_score precision_score sum topk size statistic confusion_matrix t eq mul_ expand_as append numpy max eval AverageMeter time join str copyfile model_dir save print str param_groups topk size t eq mul_ expand_as append sum max items list print copy_ from_numpy state_dict load open _read_path_label list print change_emotion_label_same_as_affectnet shuffle dict append keys range len int print readlines close change_emotion_label_of_expw_same_as_affectnet open append split int print close dict append range open size unfold unsqueeze permute append regionnet size rand extract_patches_attentivefeatures device net exp sigmoid abs load_url ResNet load_state_dict load_url ResNet load_state_dict load_url ResNet load_state_dict load_url ResNet load_state_dict load_url ResNet load_state_dict model

# ABAW2020 DMACS SSSIHL <strong>This is code for our submission in the expression track of ABAW 2020 competition.</strong> <strong> Results: https://ibug.doc.ic.ac.uk/resources/fg-2020-competition-affective-behavior-analysis/ </strong> Link to Presentation in FG-2020: https://drive.google.com/file/d/1loxCTklHu5hhkA_3pq7qIGrmZZ7HoP_y/view?usp=sharing <strong>Title</strong>: Affect expression behaviour analysis in the wild using spatio-channel attention and Complementary Context Information Paper: http://arxiv.org/abs/2009.14440 <strong> Our proposed FER framework:</strong>  <strong>Spatio-Channel Attention Network(SCAN) :</strong>

1980x/SCAN-CCI-FER

['facial expression recognition']

['Landmark Guidance Independent Spatio-channel Attention and Complementary Context Information based Facial Expression Recognition', 'Affect Expression Behaviour Analysis in the Wild using Spatio-Channel Attention and Complementary Context Information']

GACNN/main_sfew.py dataset/ckplus_dataset_cv.py dataset/affectnet_dataset.py main_sfew.py main_oulucasia.py dataset/sampler.py GACNN/main_fplus.py utils/util.py OADN/models/resnet.py dataset/ferplus_dataset.py GACNN/ferplus_dataset.py OADN/train_ferplus.py OADN/dataset/rafdb_dataset_attentionmaps.py OADN/train_sfew.py OADN/dataset/sfew_dataset_attentionmaps.py GACNN/affectnet_dataset.py OADN/train_affectnet.py GACNN/main_jaffe.py GACNN/model.py OADN/train_rafdb.py OADN/dataset/affectnet_dataset_attentionmaps.py OADN/util.py OADN/models/attentionnet.py main_ferplus.py main_ckplus.py OADN/dataset/sampler.py main_affectnet_rafdb_test_fedro.py main_affectnet.py dataset/fedro_dataset.py models/attentionnet.py dataset/rafdb_dataset.py GACNN/jaffe_dataset.py main_rafdb.py GACNN/main_affectnet.py dataset/oulucasia_dataset_cv.py dataset/sfew_dataset.py OADN/dataset/ferplus_dataset_attentionmaps.py GACNN/sfew_dataset.py dataset/affectnet_rafdb_dataset.py models/resnet.py validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy adjust_learning_rate save_checkpoint main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy adjust_learning_rate save_checkpoint main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train switch_expression ImageList default_reader PIL_loader get_class PIL_loader change_emotion_label_same_as_affectnet default_reader_affectnet ImageList default_reader_rafdb switch_expression get_class PIL_loader loademotionlabels change_emotion_label_same_as_affectnet default_reader ImageList ImageList map_emotion_label default_reader PIL_loader PIL_loader _process_data default_reader ImageList make_emotion_compatible_to_affectnet get_class PIL_loader change_emotion_label_same_as_affectnet default_reader ImageList get_class PIL_loader change_emotion_label_same_as_affectnet default_reader ImageList ImbalancedDatasetSampler ImageList default_reader PIL_loader get_class PIL_loader average_point convert68to24 _process_data default_reader ImageList make_emotion_compatible_to_affectnet get_class PIL_loader convert68to24 change_emotion_label_same_as_affectnet default_reader ImageList validate AverageMeter accuracy ImbalancedDatasetSampler save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train Net View get_class PIL_loader convert68to24 change_emotion_label_same_as_affectnet default_reader ImageList RegionBranch eca_layer Attentiomasknet count_parameters extract_patches_attentivefeatures SharedAttentionBranch AttentionBranch main norm_angle count_parameters ResNet resnet50 Bottleneck resnet152 sigmoid conv3x3 resnet34 resnet18 main BasicBlock resnet101 validate AverageMeter accuracy ImbalancedDatasetSampler save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train validate AverageMeter accuracy save_checkpoint adjust_learning_rate main train accuracy AverageMeter load_state_dict get_class PIL_loader convert68to24 _process_data _gaussian draw_gaussian default_reader ImageList make_emotion_compatible_to_affectnet get_class PIL_loader convert68to24 change_emotion_label_same_as_affectnet _gaussian draw_gaussian default_reader ImageList ImbalancedDatasetSampler get_class PIL_loader convert68to24 change_emotion_label_same_as_affectnet _gaussian draw_gaussian default_reader ImageList LandmarksAttentionBranch RegionBranch count_parameters main norm_angle count_parameters ResNet resnet50 Bottleneck resnet152 sigmoid conv3x3 resnet34 resnet18 main BasicBlock resnet101 load_state_dict validate batch_size count_parameters warn SGD pretrained DataLoader save_checkpoint dataset max str ImbalancedDatasetSampler load_state_dict AttentionBranch parse_args to ImageList sum range format resnet50 Compose start_epoch add_param_group resume item power get_cls_num_list load RegionBranch print imagesize isfile epochs array len zero_grad region_model to range cat update format size mean item basemodel enumerate time attention_model criterion backward print AverageMeter accuracy step len eval AverageMeter time join model_dir save param_groups topk size t eq mul_ expand_as append sum max adjust_learning_rate TrainList TestList train folds replace str int readlines close shuffle strip dict split append range open int format get_class print change_emotion_label_same_as_affectnet readlines close open append sum range split int print close append open dict loademotionlabels map_emotion_label sum max range len replace list set change_emotion_label_same_as_affectnet asarray average_point astype float32 int32 append zeros len items enumerate int reshape transpose array model copyfile print str state_dict print convert68to24 size unfold unsqueeze permute append regionnet size rand extract_patches_attentivefeatures device net sigmoid abs load_url ResNet load_state_dict load_url ResNet load_state_dict load_url ResNet load_state_dict load_url ResNet load_state_dict load_url ResNet load_state_dict model num_classes LandmarksAttentionBranch max items list print copy_ from_numpy state_dict exp empty range sum _gaussian format get_class sum

Our work has been published in Pattern Recognition letters. https://authors.elsevier.com/a/1ca-7cAmylz0f https://www.sciencedirect.com/science/article/abs/pii/S0167865521000489#absh0002 <strong>Title:</strong> Landmark Guidance Independent Spatio-channel Attention and Complementary Context Information based Facial Expression Recognition https://arxiv.org/pdf/2007.10298v1.pdf This work was presented in FG-2020 workshop: Affect Recognition in-the-wild: Uni/Multi-Modal Analysis & VA-AU-Expression Challenges IEEE Face and Gesture Recognition (FG-2020) Competition: Affective Behavior Analysis in-the-wild (ABAW) Leaderboard: https://ibug.doc.ic.ac.uk/resources/fg-2020-competition-affect Darshan Gera and S. Balasubramanian."Affect Expression Behaviour Analysis in the Wild (ABAW) using Spatio-channel Attention and Complementary Context Information", Arxiv preprint: http://arxiv.org/abs/2009.14440 Source Code: https://github.com/1980x/ABAW2020DMACS

198808xc/OrganSegC2F

['pancreas segmentation']

['A Fixed-Point Model for Pancreas Segmentation in Abdominal CT Scans']

OrganSegC2F/init.py OrganSegC2F/DataF.py OrganSegC2F/utils.py OrganSegC2F/fine_surgery.py OrganSegC2F/coarse_training.py OrganSegC2F/coarse_surgery.py OrganSegC2F/oracle_testing.py OrganSegC2F/coarse_fusion.py DATA2NPY/nii2npy.py OrganSegC2F/coarse_testing.py DATA2NPY/dicom2npy.py OrganSegC2F/fine_training.py OrganSegC2F/oracle_fusion.py OrganSegC2F/DataC.py OrganSegC2F/coarse2fine_testing.py transplant upsample_filt expand_score interp DataLayer DataLayer transplant upsample_filt expand_score interp snapshot_name_from_timestamp training_set_filename volume_filename_testing result_name_from_timestamp result_name_from_timestamp_s is_organ snapshot_name_from_timestamp_s testing_set_filename volume_filename_coarse2fine volume_filename_fusion post_processing snapshot_filename log_filename DSC_computation in_training_set print params range flat len data num upsample_filt print min shape range max join str str join sort len reversed listdir range snapshot_filename str join volume_filename_testing sort reversed listdir range len sum zeros_like where nonzero zeros sum array range

# OrganSegC2F: a coarse-to-fine organ segmentation framework version 1.11 - Dec 3 2017 - by Yuyin Zhou and Lingxi Xie ### Please note: an improved version of OrganSegC2F named OrganSegRSTN is available: https://github.com/198808xc/OrganSegRSTN It outperforms OrganSegC2F by 84.50% vs. 82.37% on the NIH pancreas segmentation dataset. #### Also NOTE: some functions have been optimized in the repository OrganSegRSTN, but not yet been transferred here. I will do these things in the near future - they do not impact performance, but will make the testing process MUCH faster. #### **Yuyin Zhou is the main contributor to this repository.** Yuyin Zhou proposed the algorithm, created the framework and implemented main functions. Lingxi Xie later wrapped up these codes for release. #### If you use our codes, please cite our paper accordingly:

198808xc/OrganSegRSTN

['pancreas segmentation']

['Recurrent Saliency Transformation Network: Incorporating Multi-Stage Visual Cues for Small Organ Segmentation', 'A Fixed-Point Model for Pancreas Segmentation in Abdominal CT Scans']

OrganSegRSTN/surgery.py OrganSegRSTN/Crop.py OrganSegRSTN/coarse2fine_testing.py OrganSegRSTN/oracle_fusion.py OrganSegRSTN/Uncrop.py OrganSegRSTN/coarse_fusion.py OrganSegRSTN/Crop_old.py OrganSegRSTN/fast_functions.py OrganSegRSTN/indiv_training.py OrganSegRSTN/init.py OrganSegRSTN/coarse_testing.py OrganSegRSTN/Data.py OrganSegRSTN/joint_training.py OrganSegRSTN/oracle_testing.py OrganSegRSTN/utils.py DATA2NPY/dicom2npy.py DATA2NPY/nii2npy.py CropLayer CropLayer DataLayer _swig_repr swig_import_helper _swig_setattr_nondynamic _swig_getattr post_processing _swig_setattr DSC_computation transplant upsample_filt expand_score interp UncropLayer snapshot_name_from_timestamp volume_filename_testing snapshot_name_from_timestamp_2_s result_name_from_timestamp_s is_organ snapshot_filename in_training_set snapshot_name_from_timestamp_2 testing_set_filename result_name_from_timestamp_2_s log_filename result_name_from_timestamp valid_loss training_set_filename result_name_from_timestamp_2 snapshot_name_from_timestamp_s volume_filename_coarse2fine volume_filename_fusion post_processing DSC_computation find_module load_module get get __repr__ print params range flat len data num print shape upsample_filt max join str int min len splitlines float range find str join sort len reversed listdir range snapshot_filename str join sort len reversed listdir range snapshot_filename str join volume_filename_testing sort reversed listdir range len str join volume_filename_testing sort reversed listdir range len zeros

# OrganSegRSTN: an end-to-end coarse-to-fine organ segmentation framework version 2.0 - Jul 31 2018 - by Qihang Yu, Yuyin Zhou and Lingxi Xie ### NOTEs: #### 1. v2.0 is a MAJOR update to v1.0, which we: (1) slightly changed network architecture (score layers are removed and change to a saliency layer), so that network training is more robust (especially on some tiny targets such as pancreatic cysts); (2) carefully optimized codes so that the testing stage becomes much more efficient, especially when you use multiple processes to run different folds or datasets; (3) re-implemented two functions "post-processing" and "DSC_computation" in C, which is much faster. Note that our pre-trained models are also updated.

1989Ryan/Semantic_SLAM

['semantic slam', 'semantic segmentation', 'autonomous driving']

['Visual Semantic SLAM with Landmarks for Large-Scale Outdoor Environment']

Third_Part/PSPNet_Keras_tensorflow/train.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/convert.py Third_Part/PSPNet_Keras_tensorflow/ade20k_labels.py Third_Part/PSPNet_Keras_tensorflow/python_utils/callbacks.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/caffenet.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/__init__.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/dataset.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/__init__.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/transformers.py Third_Part/PSPNet_Keras_tensorflow/Semantic_Information_Publisher.py Third_Part/PSPNet_Keras_tensorflow/drawImage/drawModule.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/tensorflow/network.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/mnist/finetune_mnist.py Third_Part/PSPNet_Keras_tensorflow/python_utils/utils.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/helper.py Third_Part/PSPNet_Keras_tensorflow/layers_builder.py Third_Part/PSPNet_Keras_tensorflow/pascal_voc_labels.py Third_Part/PSPNet_Keras_tensorflow/weight_converter.py Third_Part/PSPNet_Keras_tensorflow/cityscapes_labels.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/tensorflow/transformer.py Third_Part/PSPNet_Keras_tensorflow/python_utils/preprocessing.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/vgg.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/resnet.py src/cluster.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/alexnet.py Third_Part/PSPNet_Keras_tensorflow/Semantic_Information_Publisher_C.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/layers.py src/DBSCAN.py src/map_engine.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/caffe/__init__.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/shapes.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/errors.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/caffe/caffepb.py Third_Part/PSPNet_Keras_tensorflow/drawImage/__init__.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/graph.py src/nearbyGPS.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/tensorflow/__init__.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/classify.py Third_Part/PSPNet_Keras_tensorflow/pspnet.py catkin_ws/.ycm_extra_conf.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/kaffe/caffe/resolver.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/googlenet.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/validate.py Third_Part/PSPNet_Keras_tensorflow/Image_converter.py Third_Part/PSPNet_Keras_tensorflow/caffe-tensorflow/examples/imagenet/models/nin.py GetCompilationInfoForFile IsHeaderFile MakeRelativePathsInFlagsAbsolute FlagsForFile DirectoryOfThisScript cluster expand map_engine temp_read NearbySearch Transform GoogleMaps import_labels_from_mat assureSingleInstanceName Image_converter residual_conv short_convolution_branch BN empty_branch ResNet residual_short interp_block residual_empty build_pyramid_pooling_module Interp build_pspnet generate_color_map generate_voc_labels PSPNet PSPNet50 PSPNet101 Semantic_Imformation_Publisher Semantic_Imformation_Publisher PSPNet train set_npy_weights rot90 main validate_arguments convert fatal_error main classify display_results process_image ImageProducer ImageNetProducer main validate load_model AlexNet CaffeNet GoogleNet alexnet_spec DataSpec std_spec get_models get_data_spec NiN ResNet50 ResNet101 ResNet152 VGG16 gen_data_batch gen_data print_stderr KaffeError Graph Node GraphBuilder NodeMapper NodeKind NodeDispatchError NodeDispatch LayerAdapter shape_data shape_not_implemented get_filter_output_shape shape_concat shape_convolution shape_inner_product shape_scalar shape_pool get_strided_kernel_output_shape shape_mem_data shape_identity SubNodeFuser BatchNormScaleBiasFuser DataReshaper BatchNormPreprocessor ReLUFuser DataInjector ParameterNamer NodeRenamer ReductionParameter HingeLossParameter BlobProto BlobProtoVector NetStateRule LayerParameter PowerParameter FillerParameter ArgMaxParameter V0LayerParameter InnerProductParameter ConvolutionParameter SolverState EltwiseParameter LossParameter SliceParameter BatchNormParameter WindowDataParameter DummyDataParameter HDF5OutputParameter TanHParameter TransformationParameter SoftmaxParameter ConcatParameter DataParameter SPPParameter ParamSpec EmbedParameter SolverParameter InputParameter MVNParameter ContrastiveLossParameter NetState NetParameter BiasParameter CropParameter DropoutParameter PoolingParameter Datum SigmoidParameter BlobShape ExpParameter AccuracyParameter LogParameter ThresholdParameter TileParameter MemoryDataParameter LRNParameter ReLUParameter ImageDataParameter ELUParameter ReshapeParameter InfogainLossParameter ScaleParameter V1LayerParameter HDF5DataParameter PReLUParameter FlattenParameter PythonParameter show_fallback_warning CaffeResolver has_pycaffe get_caffe_resolver layer Network MaybeActivated TensorFlowNode get_padding_type TensorFlowEmitter TensorFlowTransformer TensorFlowMapper BaseDraw LrReducer callbacks data_generator_s31 update_inputs generate preprocess_img to_color debug print_activation class_image_to_image color_class_image add_color array_to_str append join startswith IsHeaderFile compiler_flags_ exists compiler_flags_ GetCompilationInfoForFile compiler_working_dir_ MakeRelativePathsInFlagsAbsolute DirectoryOfThisScript remove strip readlines close open append split Label print append loadmat range str str residual_conv short_convolution_branch residual_conv empty_branch residual_empty residual_short print range print exit print tuple interp_block print ResNet SGD Model build_pyramid_pooling_module Input compile zeros bitget array range generate_color_map print enumerate join layers print name reshape item set_weights join load_model weights fit_generator set_npy_weights data_generator_s31 listdir build_pspnet len range print_stderr exit fatal_error print_stderr TensorFlowTransformer transform_data def_path add_argument convert code_output_path validate_arguments caffemodel ArgumentParser parse_args data_output_path phase basename format print round argmax enumerate get_data_spec float32 placeholder ImageProducer GoogleNet model_path image_paths classify to_float pack resize_images minimum slice to_int32 format print float32 placeholder get_models get_data_spec __name__ get_output format print placeholder int32 get_data_spec float in_top_k len validate load_model model exit ImageNetProducer get_data_spec list reshape shuffle images range len append next range gen_data write pad_h stride_w pad_w kernel_h kernel_w float stride_h height hasattr get_filter_output_shape kernel_parameters output_shape parameters width output_shape parameters list parents axis output_shape output_shape CaffeResolver write ceil float height width ReduceLROnPlateau ModelCheckpoint TensorBoard astype imresize join list defaultdict print shuffle listdir values join imresize update_inputs zoom astype shuffle imread enumerate color zeros uint8 range class_image_to_image add_color shape range zeros to_color print_activation print Model predict

# Semantic SLAM  This on-going project is Semantic SLAM using ROS, ORB SLAM and PSPNet101. It will be used in autonomous robotics for semantic understanding and navigation. Now the visualized semantic map with topological information is reachable, where yellow represents buildings and constructions, green represents vegetation, blue represents vehicles, and red represents roads and sidewalks. The cube is ambiguous building location and green line is the trajectory. You can visualize these information using Rviz.  You can also get the semantic topological map which only contains the ambiguous building location and trajectory. The whole ROS communication structure of the project is shown below.  ## Bibliograhpy If you are going to use our work in your research, please use the citation below.

1Reinier/Reservoir

['time series']

['Efficient Optimization of Echo State Networks for Time Series Datasets']

docs/conf.py reservoir/detail/robustgpmodel.py reservoir/esn.py setup.py reservoir/esn_cv.py reservoir/scr.py tests/test_esn.py reservoir/__init__.py reservoir/clustering.py reservoir/detail/esn_bo.py setup_package ClusteringBO EchoStateNetwork EchoStateNetworkCV generate_states_inner_loop SimpleCycleReservoir EchoStateBO FixedInverseGamma RobustGPModel Suppressor test_esn argv setup intersection zeros hstack range tanh show optimize EchoStateNetworkCV plot print loadtxt reshape train test title legend EchoStateNetwork predict

Reservoir ========= A Python 3 toolset for creating and optimizing Echo State Networks. >Author: Jacob Reinier Maat, Nikos Gianniotis >License: MIT >2016-2019 Contains: - Vanilla ESN and Simple Cyclic Reservoir architectures. - Bayesian Optimization with optimized routines for Echo State Nets through `GPy`. - Clustering routines to cluister time series by optimized model.

201518018629031/HGATRD

['graph attention']

['Heterogeneous Graph Attention Networks for Early Detection of Rumors on Twitter']

train.py models.py gat.py layers.py utils.py gcn.py SpGAT GCN GraphConvolution SpGraphAttentionLayer SpecialSpmm SpecialSpmmFunction Model evaluate pass_data_iteratively test adjust_learning_rate train evaluation_4class normalize_adj sparse_mx_to_torch_sparse_tensor load_user_tweet_graph accuracy build_symmetric_adjacency_matrix load_data normalize convert_to_one_hot load_vocab_len sum format arange LongTensor batch_size model backward nll_loss evaluate zero_grad print numpy item step range len arange len eval append range detach param_groups elapsed_time format tweets_count print pass_data_iteratively tolist dataset classification_report save accuracy_score state_dict evaluation_4class print classification_report pass_data_iteratively tolist cuda format normalize_adj LongTensor print tuple multiply sparse_mx_to_torch_sparse_tensor shape eye info range len diags flatten dot sum array sum type_as double data Size astype float32 from_numpy shape int64 format normalize_adj LongTensor print tuple multiply sparse_mx_to_torch_sparse_tensor eye range len flatten sum array diags multiply coo_matrix normalize_adj eye str index round float numpy max range len

# HGATRD The implementation of our IJCNN 2020 paper "Heterogeneous Graph Attention Networks for Early Detection of Rumors on Twitter" # Requirements python 3.6.6 numpy==1.17.2 scipy==1.3.1 pytorch==1.1.0 scikit-learn==0.21.3 # How to use ## Dataset

21lva/EEN_Tensorflow

['video prediction']

['Prediction Under Uncertainty with Error-Encoding Networks']

visualize.py new_train_d.py new_models.py new_train_s.py utils.py dataloaders/data_atari.py LatentNetwork DeterministicNetwork validation_epoch train train_epoch validation_epoch train train_epoch read_data log load_model ImageLoader get_batch print epoch_size apply_gradients compute_gradients range compute_gradients epoch_size get_batch range format model nEpoch system train_epoch datapath log save_weights filename append range validation_epoch open str print write system now close dirname open get eval cpu

# EEN_Tensorflow error encoding network by tensorflow only for breakout. The paper of EEN : https://arxiv.org/abs/1711.04994

21lva/EEN_acrobot

['video prediction']

['Prediction Under Uncertainty with Error-Encoding Networks']

train_d.py expert_policies/utils/tools.py pa.py models.py predict.py utilsf.py train_s.py changeJson.py LatentNetwork DeterministicNetwork DataMaker load_model DrawGraph invScaling BaseLineModelPredictor mape get_cond LatentModelPredictor validation_epoch train train_epoch validation_epoch train train_epoch read_data log json_to_par make_dir_if_none parse_input_str str2bool LatentNetwork DeterministicNetwork AdamOptimizer load_weights lr ri get_batch range model nState clf open str title ylim savefig append range plot npred close tfilename nAction ncond reader nTransporter print system DrawGraph model compute_loss get_cond save_dir datapath open writer tolist filename range predict npred close mean tfilename ncond nTransporter print writerow system decode DrawGraph model mae ri compute_loss get_cond save_dir datapath open writer invScaling tolist mape filename append range convert_to_tensor format npred close tfilename ncond print writerow system float32 mean_squared_error norm get_batch epoch_size mean apply_gradients compute_gradients range get_batch epoch_size mean compute_gradients range model nEpoch save_weights clf validation_epoch log datapath writer open title savefig filename append range format plot close writerow system train_epoch figure loss open __setattr__ list read_data keys str print write system now close dirname open parse_args add_argument ArgumentParser print exists makedirs

### EEN_acrobot Error encoding network for acrobot error encoding network by tensorflow.keras. The paper of EEN : https://arxiv.org/abs/1711.04994 need: python>=3.5 tensorflow (any version that support tf.keras) numpy matplotlib etc....

3778/Ward2ICU

['time series']

['Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs', 'Ward2ICU: A Vital Signs Dataset of Inpatients from the General Ward']

tests/test_samplers.py run-experiment.py tests/test_models.py ward2icu/models/classifiers.py ward2icu/metrics.py ward2icu/models/cnngan.py ward2icu/models/__init__.py ward2icu/trainers.py ward2icu/samplers.py tests/test_data.py ward2icu/models/rgan.py ward2icu/utils.py ward2icu/models/rcgan.py tests/test_utils.py tests/test_trainers.py ward2icu/layers.py ward2icu/__init__.py main cli test_TimeSeriesVitalSigns_transform_minmax_signals test_TimeSeriesVitalSigns_transform_minmax test_TimeSeriesVitalSigns_inv_transforms test_RCGAN test_BinaryRNNClassifier test_RGAN_forward test_FCCMLPGAN RGAN_setup test_BinaryCNNClassifier test_BinaryBalancedSampler test_BinaryBalancedSampler_batch test_IdentitySampler test_MinMaxGANTrainer_calc_gen_loss test_BinaryClassificationTrainer_train binaryclass_tiled_trainer test_MinMaxGANTrainer_calc_dsc_loss test_MinMaxGANTrainer_train minmaxgan_trainer binaryclass_trainer test_MinMaxGANTrainer_calculate_metrics test_BinaryClassificationTrainer_calculate_metrics_tiled test_BinaryClassificationTrainer_calculate_metrics test_tile Conv1dLayers AppendEmbedding rnn_layer Permutation View mean_feature_error classify tstr IdentitySampler BinaryBalancedSampler BinaryClassificationTrainer numpy_to_cuda flatten tile calc_conv_output_length n_ BinaryRNNClassifier BinaryCNNClassifier CNNCGANGenerator CNNCGANDiscriminator RCGANGenerator RCGANDiscriminator RGANDiscriminator RGANGenerator main cli test_TimeSeriesVitalSigns_transform_minmax_signals test_TimeSeriesVitalSigns_transform_minmax test_TimeSeriesVitalSigns_inv_transforms test_RCGAN test_BinaryRNNClassifier test_RGAN_forward test_FCCMLPGAN RGAN_setup test_BinaryCNNClassifier test_BinaryBalancedSampler test_BinaryBalancedSampler_batch test_IdentitySampler test_MinMaxGANTrainer_calc_gen_loss test_BinaryClassificationTrainer_train binaryclass_tiled_trainer test_MinMaxGANTrainer_calc_dsc_loss test_MinMaxGANTrainer_train minmaxgan_trainer binaryclass_trainer test_MinMaxGANTrainer_calculate_metrics test_BinaryClassificationTrainer_calculate_metrics_tiled test_BinaryClassificationTrainer_calculate_metrics test_tile Conv1dLayers AppendEmbedding rnn_layer Permutation View mean_feature_error classify tstr IdentitySampler BinaryBalancedSampler BinaryClassificationTrainer numpy_to_cuda flatten tile calc_conv_output_length n_ BinaryRNNClassifier BinaryCNNClassifier CNNCGANGenerator CNNCGANDiscriminator RCGANGenerator RCGANDiscriminator RGANDiscriminator RGANGenerator main generator model div abs cuda SequenceTrainer BinaryBalancedSampler tstr discriminator classify synthesis_df log_model trainer size info sample log_to_mlflow log_metric set_tag log_df TimeSeriesVitalSigns mean_feature_error numpy float128 synthesis_df y astype array array dict RGANDiscriminator RGANGenerator forward sampler RCGANGenerator RCGANDiscriminator sampler dict forward len BinaryRNNClassifier forward randn forward BinaryCNNClassifier randn sampler dict FCCMLPGANDiscriminator forward FCCMLPGANGenerator len BinaryBalancedSampler eye sample Tensor range BinaryBalancedSampler eye sample Tensor range ones Tensor IdentitySampler sample IdentitySampler SGD parameters cpu float BinaryClassificationTrainer IdentitySampler SGD parameters cpu float BinaryClassificationTrainer MinMaxBinaryCGANTrainer RCGANGenerator randn IdentitySampler RCGANDiscriminator SGD parameters randint train mean Tensor calc_dsc_real_loss calc_dsc_fake_loss calc_gen_loss Tensor Tensor calculate_metrics train sigmoid Tensor sum calculate_metrics sigmoid T sum calculate_metrics Tensor tile dict numpy_to_cuda BinaryBalancedSampler IdentitySampler Adam parameters info train cuda BinaryClassificationTrainer numpy_to_cuda BinaryBalancedSampler IdentitySampler train_test_split_tensor Adam parameters info train cuda BinaryClassificationTrainer _maybe_slice padding kernel_size stride dilation

--- <div align="center"> # Ward2ICU [](https://arxiv.org/abs/1910.00752) [](https://research.3778.care/projects/privacy/) [](https://forms.gle/e2asYSVaiuPUUCKu8) </div> <!--ts--> * [Description](#description)

3lis/rnn_vae

['autonomous driving']

['On the Road with 16 Neurons: Mental Imagery with Bio-inspired Deep Neural Networks']

src/tester.py src/exec_eval.py src/gener.py src/h5lib.py src/exec_main.py src/losses.py src/cnfg.py src/exec_lata.py src/arch.py src/exec_feat.py src/exec_dset.py src/sample_sel.py src/trainer.py src/mesg.py src/pred.py RecMTimepred save_model Rec2Timepred RecTimepred RecMultiAE RecMultiVAE load_model load_vgg16 RecAE VAE AE get_unb_loss jaccard_loss MultipleVAE get_loss create_model stanh model_summary Autoencoder Multiple get_optim pre_unb_ones RecVAE Timepred MultipleAE Recursive get_config load_config get_args get_args_eval sel_test dset_class get_files_noclass dset_seq convert_gt dset_sel get_pos dset_time plain_link_synthia get_odom resize_img set_mseq_tree dset_seq_multi next_file save_dset_time get_files_class dict_odom next_timeframe check_valid dset_noclass frac_dataset resize_synthia make_symlink read_pose get_pitch load_dset_time check_timeseq pred_models frmt_result write_header eval_model_set eval_model pred_set recreate_model get_multi_decoders pred_model do_inspect_model pred_time get_decoder recall_img inspect_model eval_models get_encoder gen_latent dset_latent base_stat plot_seq_stat get_args seq_stat plot_code_stat code_stat read_latent test_model init_config archive create_model dset_class len_dataset iter_simple dset_multiple iter_multi_seq iter_double iter_seq iter_multiple dset_sequence dset_same gen_dataset dset_multi_seq load_h5_layer load_h5_group valid_layer load_h5 invalid_layer h5_multi_groups h5_single_group load_h5_vgg16 xentropy unb_lane array_to_image mse save_loss get_unb_pars unb_car get_unb_par image_to_array print_err print_wrn print_line print_msg TimeInterPred eval_latent in_rgb pred_tset eval_img create_test_vae accur_img model_dead accur_mutliseq get_targ eval_tset img_overlays pred_folder save_image interpolate_pred code_to_latent get_metric img_overlay eval_multitset file_to_array pred_image accur_sel_time model_outputs hallucinate pred_rec_tset model_weights eval_latent_folder create_test_model pred_time_sample swap_latent accur_tset_time create_test_mvae layer_outputs array_to_image pred_multi_image pred_multirec_image eval_multimg pred_time latent_to_image pred_multi_tset eval_raw pred_multirec_tset accur_class_latent code_to_path save_collage make_collage imgsize accur_tset save_gif layer_weights accur_class_multiseq out_rgb accur_latent rearrange_rtime ChangeKlWeight plot_history set_callback generate_dset_time train_ae_model train_time_model print_err format join lower load_img img_to_array pre_unb_ones abs sum print_err format join plot_model print_summary makedirs endswith join get_file join getcwd makedirs join save_weights RecVAE RecMTimepred format Timepred Rec2Timepred MultipleAE print_err RecAE VAE RecTimepred RecMultiAE RecMultiVAE AE MultipleVAE compile add_argument ArgumentParser add_argument ArgumentParser format print_msg print_err system dict list keys append format append sorted next_timeframe range str dirname len reshape array swapaxes split norm format print exit next_file read_pose get_pos get_pitch join get_odom format print len int join next_timeframe sorted format print reshape next_timeframe File close set array append keys range seed join list format print save frac_dataset sample dset_time range len load join slice int format print join isinstance relpath symlink isfile range len close array open ANTIALIAS size min convert save resize crop open fromarray read format list uint8 print exit Reader save array join resize_img format print rmtree convert_gt exists makedirs join format replace print rmtree func exists makedirs sorted format isinstance print append listdir sorted format isinstance print append listdir range len seed join get_files_noclass make_symlink rmtree frac_dataset sample exists makedirs seed list len make_symlink rmtree frac_dataset sample range array exists get_files_class makedirs seed list format get_files_noclass next_timeframe makedirs make_symlink sample rmtree array frac_dataset append range check_timeseq exists enumerate len format rmtree append exists makedirs seed set_mseq_tree list next_timeframe make_symlink sample array frac_dataset append range check_timeseq enumerate get_files_class len center format write expandtabs len max list std mean array values join list format create_model dset_time_neat print get_config exit dset_time keys compile print name format exit print name format exit format layers print name exit join eval_multitset ndarray isinstance eval_tset mean lower join recreate_model eval_model_set clear_session join write_header eval_model_set recreate_model write close open model_outputs model_weights create_test_model clear_session join recreate_model isfile inspect_model makedirs recreate_model get_decoder join lower pred_tset clear_session join pred_set recreate_model pred_time isfile makedirs clear_session join recreate_model pred_model clear_session join format print pred_image recreate_model exit isfile out_rgb save_image print join pred_image walk join list format print File close gen_latent create_dataset keys get_encoder makedirs join sorted format list print reshape File exit close shape array keys mean std base_stat clip append len base_stat format append clip len set_aspect arange len close pcolormesh savefig yticks set_aspect arange len close pcolormesh savefig yticks join format set_session isinstance print_err makedirs len strftime set_random_seed eval ConfigProto Session open join save_model model plot_history train_ae_model train_time_model clear_session pred_tset model eval_tset generate_dset_time interpolate_pred pred_folder eval_multitset time_test_set print_wrn recreate_model accur_sel_time hallucinate pred_rec_tset eval_latent_folder range get_encoder nn_best pred_time_sample format create_model accur_tset_time lower pred_time pred_multi_tset TimeInterPred pred_multirec_tset enumerate join accur_class_latent isinstance print makedirs accur_tset time_test_str get_decoder accur_class_multiseq len join format system makedirs ImageDataGenerator flow_from_directory format iter_simple save_image next range format iter_simple append save_image next range len format iter_simple copy append save_image next range enumerate len format iter_simple append save_image range enumerate len join iter_simple join iter_double join iter_multiple join iter_seq iter_multi_seq len walk len print_err format format print_wrn name set_weights shape get_weights array len list format print_wrn get_layer keys enumerate invalid_layer print_wrn format enumerate list File valid_layer keys list format print_wrn name File search valid_layer keys shape get_weights array enumerate set_weights join close float array open get_unb_par join zeros_like clip log where get_unb_pars xentropy get_unb_pars xentropy fromarray uint8 ptp min convert mean expand_dims close array open array_to_image save f_back basename format write exit getfile f_lineno f_back basename format write getfile flush f_lineno write isinstance input_shape isinstance isinstance output_shape expand_dims img_to_array load_img split_latent ndarray isinstance variable array_to_image eval img_overlays out_rgb expand_dims predict load join int array_to_image ndarray isinstance save new paste resize range len save make_collage save size new convert array_to_image where filter FIND_EDGES composite img_overlay ndarray isinstance array_to_image save array print_err format replace file_to_array in_rgb file_to_array isinstance out_rgb save_image predict in_rgb file_to_array isinstance img_overlays save_image predict in_rgb format isinstance makedirs enumerate img_overlays append save_image range predict len join replace pred_image pred_multi_image img_overlays save open makedirs join list sorted save_collage imgsize linspace out_rgb array range len join sorted list save_collage imgsize open linspace append array range len join sorted list save_collage size array_to_image linspace zip append array range len join sorted list save_collage size img_overlays linspace zip append array range len linspace list map append range predict format glob latent_to_image swapaxes zip join save_collage isinstance print_err imgsize out_rgb array len int join format list replace save_collage makedirs len imgsize save_gif save zip append expand_dims array range predict open save open list len append expand_dims range predict format replace update_frame latent_to_image zip int join save_collage imgsize save_gif array makedirs in_rgb variable img_overlays linspace save open split_latent file_to_array append predict format replace insert eval enumerate join save_collage imgsize save_gif makedirs join in_rgb split_latent file_to_array list save_collage concatenate variable map imgsize eval img_overlays save open predict makedirs logical_or sum logical_and logical_not append accur_img predict zip accur_img latent_to_image zip append expand_dims array predict items sorted file_to_array format list accur_img pred_image write close nanmean zip out_rgb array values open items sorted format list values print print_err write close nanmean open append accur_latent array range flush len sel_test values open sorted list header check_incl append range format next_timeframe close flush items print write nanmean accur_latent array sel_test accur_mutliseq values open sorted list header check_incl append range format next_timeframe close flush items print write nanmean array sel_test code_to_path format print accur_img write get_targ close nanmean latent_to_image check_incl append code_to_latent flush open latent_to_image code_to_latent get_targ eval_latent join save makedirs out_rgb in_rgb file_to_array evaluate file_to_array evaluate file_to_array pred_image variable metric eval out_rgb items sorted format list eval_img values write close nanmean zip array get_metric compile open items sorted format list values tuple write close nanmean zip eval_multimg array compile open Model get_layer lower get_layer Model fromarray join format save_collage print_msg ptp min imgsize append range max join format save_collage print_msg ptp reshape min isnan append sum array range join format isinstance layer_outputs print_msg pred_image print_err imgsize zip predict layer_weights isinstance zip list format print_msg ptp map linspace zeros sum range len array swapaxes rearrange_rtime load_dset_time join TensorBoard EarlyStopping ChangeKlWeight ceil ModelCheckpoint append makedirs len_dataset multi_gpu_model model now fit_generator gen_dataset ceil multi_gpu_model model fit now generate_dset_time len list format plot xlabel close ylabel mean ylim savefig legend std range len

# On the Road with 16 Neurons: Mental Imagery with Bio-inspired Deep Neural Networks ###### *Alice Plebe, Mauro Da Lio (2020).* --- This repository contains the source code related to the paper [*"On the Road with 16 Neurons: Mental Imagery with Bio-inspired Deep Neural Networks"*](http://arxiv.org/abs/2003.08745). The code is written in Keras using TensorFlow backend. <!-- The code is written in Keras 2.2.4 using TensorFlow 1.12.0 backend. The scripts are executed with Python 3.6.8. The networks are trained on multiple GPUs with CUDA 10.1. The neural models obtained from Keras are exported to __Wolfram Mathematica 11.3__ for visualization.

4kubo/bacf_python

['visual tracking']

['Learning Background-Aware Correlation Filters for Visual Tracking']

background_aware_correlation_filter.py special_operation/convertor.py utils/arg_parse.py special_operation/resp_newton.py utils/report.py utils/get_sequence.py image_process/feature.py utils/functions.py demo_on_otb.py BackgroundAwareCorrelationFilter show_image_with_bbox get_pixel get_pyhog get_pixels resize_DFT2 resp_newton parse_args get_subwindow_no_window get_sequence_info _get_path_to_seq _exclude load_image _get_gt_bbox _numerical_sort _get_frame_names LogManger imshow rectangle waitKey minimum astype maximum stack array minimum tuple astype maximum pad resize array float64 process astype copy squeeze min zeroes shape any array floor ceil prod int exp mod nanargmax reshape pi flatten prod floor tile real argmax max einsum add_argument ArgumentParser arange floor isscalar meshgrid array _get_path_to_seq format print _exclude _get_gt_bbox zip enumerate int format glob strip split append _get_frame_names exists enumerate len glob sorted format list map compile split array exists enumerate len append enumerate

# A port of [BACF in Matlab](http://www.hamedkiani.com/bacf.html) to python 2 Python 2 implementation of Background-Aware Correlation Filters for visual tracking. For more detail, please refer to [arXiv paper](https://arxiv.org/abs/1703.04590) and [the author's website ](http://www.hamedkiani.com/bacf.html)  # Requirements - python 2 - numpy - scipy - opencv - PIL

50kawa/mimick_chainer

['word embeddings']

['Mimicking Word Embeddings using Subword RNNs']

make_wordlist.py data_make.py util.py eval_wordsimilarity.py main.py model.py ngram zijougosa cos_sim cos_sim load_wordvector mse main calc_batch_loss cos_sim EncoderCNN load_model Interpreter Encoder EncoderGRU EncoderSumFF Model EncoderSum SplitWord Dataread PWIMdata load_word2vec_format load_fasttext_format random_cw_list extend sample cos_sim append range len data disable_update Interpreter ArgumentParser train_data GradientClipping cleargrads list use load_model setup hasattr exit Adam interpreter add_hook append parse_args range update save_hdf5 shuffle zip calc_batch_loss float enumerate check_cuda_available backward print load_wordvector add_argument test_data to_gpu len load max_input_len Model Dataread char_dict len

# mimick_chainer https://arxiv.org/abs/1707.06961 の実装をchainerでやってみました。 ## 動かし方

590shun/vsum_dsf

['video summarization']

['Video Summarization using Deep Semantic Features']

script/summarize.py func/sampling/vsum.py func/dataset/summe.py script/evaluate.py script/uniform_smpl.py func/nets/vid_enc.py func/nets/vid_enc_vgg19.py SUMME Model Model representativeness encodeSeg uniformity VSUM eval_human_summary eval_summary eval_f1 get_flabel uniform_sampling model reduce from_numpy feat sampleFrame mean getChrDistances mean getDistances int list astype load get T format eval_f1 loadmat open load get T format sum len min delete eval_f1 append loadmat max range open zeros list map zip print zeros float floor

# Video Summarization using Deep Semantic Features これは"Video Summarization using Deep Semantic Features" in ACCV'16 [[arXiv](arxiv.org/abs/1609.08758)]を書き直したものになります。 [実装元のリンク](http://github.com/mayu-ot/vsum_dsf),[備忘録](https://github.com/590shun/paper_challenge/issues/7) ## 実験手順 git clone https://github.com/590shun/vsum_dsf.git ### オプション このコードでは以下のM. Gygli *et al.* [1]を参考にしています。 環境構築は次のように行います。 cd vsum_dsf git clone https://github.com/gyglim/gm_submodular.git

5gon12eder/msc-graphstudy

['graph generation', 'data augmentation']

['Aesthetic Discrimination of Graph Layouts']

driver/quarry.py benchmarks/lib/history.py driver/www/property.py driver/www/graph.py utils/prepare-local-diffent-regression.py benchmarks/lib/__init__.py eval/progress.py driver/__init__.py eval/make-config-puncture.py utils/prepare-entropy-regression.py benchmarks/micro-driver.py driver/model.py driver/www/nn.py driver/www/perfstats.py test/driver-check.py utils/format-puncture.py driver/www/overview.py utils/format-confusion.py utils/format-huang-weights.py benchmarks/lib/manifest.py codegen/enums.py driver/cli.py driver/constants.py eval/cross-validate.py benchmarks/lib/fancy.py driver/layinter.py maintainer/maintainer.py driver/errors.py driver/crash.py driver/www/select.py report/pics/get-graph-info.py utils/download.py driver/doctests.py driver/layworse.py driver/graphs.py driver/www/property_plots.py driver/alternatives.py driver/www/__init__.py driver/manager.py driver/resources/__init__.py driver/integrity.py driver/www/badlog.py driver/compare.py benchmarks/lib/cli.py driver/www/layout.py eval/get-huang-parameters.py utils/format-competing-metrics.py utils/typeset.py driver/httpd.py driver/nn.py driver/deploy.py driver/features.py benchmarks/macro-driver.py driver/configuration.py driver/archidx.py benchmarks/history.py driver/layouts.py benchmarks/lib/runner.py driver/properties.py driver/utility.py driver/xjson.py driver/metrics.py utils/format-nn-info.py utils/picture-wrapper.py utils/format-disco.py driver/www/impl_common.py driver/imports.py driver/tools.py _action_export _action_list median_rel_stdev main is_outlier count _mean_stdev Generator MacroManifestLoader MacroCollectionRunner main _is_list_of_str MacroBenchmarkRunner MicroBenchmarkRunner main MicroCollectionRunner MicroManifestLoader _ArgInput use_color add_constraints add_argument_groups add_info add_manifest add_history_optional _ArgMaybe add_color add_help TerminalSizeHack add_alert _ArgRealFile add_benchmarks add_verbose add_history_mandatory _ArgNumber Reporter _fmttime NoAnsi _shorten Ansi _as_singleton HistoricResult _count ResultAggregation History InvalidManifestError ManifestLoader Result BenchmarkRunner CollectionRunner Constraints Failure _get_trend TimeoutFailure print_source dashy log print_header print_test main make_text_wrapper get_alternative_value get_alternative_context _get_graph_id _get_alternative_value_huang AlternativeContext _get_alternative_value_stress AlternativeContextHuang _Indexer parse_generator Main _patch_logging_module _get_log_level_environment _adjust_log_level _check_working_directory AbstractMain read_id_pairs process smart_open Main parse_dmspec make_id_output_projection get_id_pairs ValueErrorInDisguise _CfgGraphs ConfigReader _Config _CfgLayInter _CfgLayWorse _ConfigBySize _CfgImports _CfgLayouts _CfgPuncture Configuration _CfgMetrics _ConfigByRate BadLog _CfgProperties Binnings Id Layouts Tests Metrics sqlrepr LogLevels translate_enum_name Kernels enum_to_json TestStatus LayWorse _GraphSizeAttributes enum_from_json LayInter Generators Properties Actions GraphSizes dump_current_exception_trace _make_badlog parse_positive_float _perform_cleaning Main _perform_action _recursively_run_doctests _Status Error SanityError ConfigError RecoverableError FatalError _convert_result get_graph_features get_layout_features _ilog _emit_layout_features _emit_graph_features _gen_generic _quick_archive_import_eh _call_generic_tool do_graphs _BucketList _insert_layout _insert_graph _gen_import _generate_graphs _update_bucket_list RequestHandler _filename _validate_vicinity _validate_property _run_httpd_forever _ChunkedIO canonical_path Main _close_log_file _make_html_tree _open_log_file Httpd _convenient_name UrlImportSource NullImportSource DirectoryImportSource ImportSource _log_usage_notice get_well_known_import_sources TarImportSource _log_types_notice get_import_source_from_json _get_mandatory_and_optional_parameters _get_cache_file_name _check_attribute_type _isbetween _load_indexed_files Main _store_indexed_files _get_file_info Checker _interpolate_generic do_lay_inter _add_inter_layout _lay_generic do_layouts _worsen_generic _add_worse_layout do_lay_worse _log_data _has_functional_fileno_attribute _register_sqlite_types Manager _handle_popen_stdin _handle_popen_stdout _compute_metric do_metrics _get_graph_size _compute_all_metrics _insert_metric _test_alternatives _restore_features _save_test_score _do_compile_model _get_layout _check_for_non_finite_feature_vectors _DebugMain _build_model _restore_testscore _train_alternative_huang _import_3d_party_libraries _gather_worse _test_model _get_layout_features_curs _save_features _rank_layouts _gather_proper _restore_weighted_model _get_debugdir _get_normalizers _make_half_model _cache_put _dump_model_infos _train_model _log_discard_too_old _normalize_data _train_alternatives _gather_inter _cache_get _load_training_and_testing_data _DataSet _rank_layout_id _get_graph_id _graph_features _save_weighted_model _get_graph_id_curs _dump_corpus_infos _get_same_graph_id _layout_features _rank_layout_ids _get_same_graph_id_curs _load_huang_context_with_matrix_and_weights_vector _setup do_model _get_sql_view_columns _rank_layout _write_test_summary _get_worse_max _get_graph_features_curs Oracle NNTestResult NNFeature do_properties _insert_data _get_directory _get_executable _sql_insert_curs _temporary_to_persistent_name _maybe_insert_pca _rename_files_batch _compute_prop_outer _compute_prop_inner _roundabout pickle_objects unpickle_objects get_cache_directory get_gnuplot_options _find_tool _get_safe_command _get_cache_directory get_gnuplot_environment find_tool_lazily find_all_tools_eagerly GetNth Real rfc5322 value_or Power singleton get_one_or prepare_fingerprint attribute_projection count get_one encoded _format_number get_first_or index_projection change_file_name_ext replace_list is_sorted get_first Override Static make_ld_library_path fmtnum Integer GetNthOr _preprocess_line XJsonError load_xjson_file load_xjson_string _load_xjson_pp get_resource_as_stream serve _fmtkey serve append_common_property get_informal_layout_name get_bool_from_query get_int_from_query append_common_all_constants get_informal_layout_name_curs append_common_graph _offer_picture_to_cache _caching_policy _NotFound _get_primary_axes_curs _get_cache_file_name _append_pca _xmlkey _check_graphics_format _serve_for_graph _serve_picture _serve_by_id serve _serve_generic _append_one_property_inner _get_primary_axis_curs _get_picture_from_cache _get_primary_axes _sqlkey _get_worse_info _append_property _get_conversion_command_from_svg _append_one_property_outer _get_inter_info _get_layout_ids_from_query _404 _500 _load_feature_rows _serve_testscore _NotFound _serve_demo _get_random_layout_pairs _serve_features _get_random_count_from_query _get_special_test_case_selection_from_query serve serve _get_perf_stats_raw _get_perf_stats_cooked serve _append_scaled_histogram_stuff _check_graphics_format _UnknownHandler _serve_sliding_average _Handler _get_sigma_from_query _GN _serve_plot _serve_histogram _get_bincount_from_query _gpltin _get_vicinities_from_query _serve_entropy serve _get_type _add_row_value link_xslt Child validate_layout_id validate_graph_id validate_id _NoXslt parse_bool HttpError format_bool append_child get_one make_statistic combine_confusions get_confusion cross_validate main combine_huang_weights Confusion main print_punctures main canonical _property store_info print_message _file _warning Info _float_nn main update_progress_report load_info Error ForEachCfg get_vpath_and_vardeps_root BuildError _BuildStatus find_build_tool pretend _print_summary Main PipeGuard report _rfc5322 looks_like_build_directory get_top_level_source_directory main format_integer write_texdef check_graph_files check_graphs b2s main check_layouts check_database_exists parse_checksum_option extract_files_tar add_to_cache invoke_argument_parser guess_archive_format get_human_rate try_download extract_files serve_from_download info_ext copy_iostreams extract_files_zip info main serve_from_cache parse_extract_option hexhash bemoan load_rename_table _test analysis_dict2tuple dummy_analysis get_lazy_okay main write_output check_lazy_okay get bemoan _test get_lazy_okay main check_lazy_okay bemoan get_lazy_okay main parse_disco_file check_lazy_okay parse_dmnamespec bemoan load_rename_table get_lazy_okay main check_lazy_okay main Parser get_layer_dim_def bemoan load_rename_table _test info_2_mean_stdev get_puncture_properties get_lazy_okay main check_lazy_okay print_output main get_princomp main main Error get_aux_checksum quote get_tool_flags save_html_report find_dependencies find_dependency run_command ap_directory prepare_dependencies get_last_mtime run_toolchain main find_executable looks_like_error_line add_argument_group add_argument add_help add_parser ArgumentParser add_history_mandatory add_subparsers sorted format get_descriptions print keys localtime description max count sorted get_benchmark_results strftime stdev printhdr format results median_rel_stdev mean n print min timestamp filter len print_info use_color load add_argument_groups Reporter print manifest Constraints color MacroManifestLoader info compile mean stdev MicroManifestLoader add_constraints add_info add_manifest add_history_optional add_argument_group add_color add_help add_alert add_benchmarks add_verbose add_argument add_argument add_argument add_argument add_argument add_argument add_argument add_argument add_argument add_argument compile sqrt join items format strip loads infile startswith parse_args log makedirs sorted format print len map keys upper fill max make_text_wrapper enumerate sorted format print dashy keys len sorted format print dashy keys enumerate print defaultdict abs _get_graph_id get_cached_mean_and_stdev append sum add set get_one_or sql_select_curs abs enum_from_json join imported join getcwd debug format get list reverse addLevelName format print Lambda map lower setattr is_alternative zip oracle read_id_pairs combinations list stdin all extend map set any append error format map enumerate WARNING ERROR CRITICAL DEBUG INFO Enum isinstance mkstemp exc_info format critical known_bad do_properties format name do_metrics now rfc5322 do_model do_layouts do_graphs do_lay_inter do_lay_worse notice clean_worse format clean_graphs name now rfc5322 clean_metrics clean_inter clean_properties clean_layouts notice clean_model float update join format print testmod write extend iter_modules perf_counter stderr timedelta import_module print_report _Status list keys frozenset puncture get_one_or sql_select_curs join sql_select_curs list items extend dict FIXED_COUNT get_one_or get_wrapper items list defaultdict desired_graphs import_sources request _generate_graphs _update_bucket_list all_items list format _quick_archive_import_eh discard_unbounded_requests name info sum max change _gen_generic format name get_well_known_import_sources imported info _gen_import len pick join format extend target abs_bindir enum_to_json append total Id make_graph_filename getenv int int format configdir Configuration isinstance int close append_child isinstance Element list filter split dict get_import_source_from_json get getfullargspec join format throw_u list items _get_mandatory_and_optional_parameters throw _check_attribute_type type notice join notice format args dict defaults reversed args get_cache_directory format startswith notice expanduser isabs split notice format notice format S_ISSOCK st_mode S_ISDIR S_ISWHT S_ISBLK S_ISDOOR stat S_ISIFO S_ISCHR S_ISLNK S_ISREG S_ISPORT _interpolate_generic list format combinations sorted name map set info LAY_INTER get_bad keys len format debug rename dirname makedirs sorted sql_select format classify name LAYOUTS _lay_generic notice get_bad enumerate len name format info _worsen_generic format list sql_select name LAY_WORSE map set info get_bad keys len format debug rename dirname makedirs format rstrip len map splitlines notice enumerate _has_functional_fileno_attribute read BytesIO isinstance DEVNULL getbuffer _has_functional_fileno_attribute PIPE DEVNULL fileno getkey register_converter register_adapter list items _compute_all_metrics sql_select format _compute_metric name info notice get_bad METRICS append call_graphstudy_tool extend _insert_metric format rstrip map info enumerate len notice format format nn_weights nn_model save_weights info format nn_weights nn_model model_from_yaml load_weights info list format zip dict nn_features NNFeature info append float format reshape nn_features info _log_discard_too_old list executemany iter append float enumerate compile tuple format sql_exec get_one_or sql_select_curs notice _get_graph_id_curs format getenv format warning nndir _import_3d_party_libraries makedirs _train_alternatives _test_alternatives _load_training_and_testing_data _setup _build_model _train_model _write_test_summary info _save_weighted_model _test_model time list format _get_normalizers reshape _DataSet _check_for_non_finite_feature_vectors set dict info _dump_corpus_infos _save_features array _normalize_data len int format info fill zeros sum combinations list sql_select_curs _rank_layouts keys sql_select_curs combinations defaultdict _get_graph_id dict keys _rank_layout_ids sql_select_curs combinations defaultdict items _rank_layout_id _get_graph_id dict _get_worse_max keys time format _get_debugdir _make_half_model _do_compile_model Model _dump_model_infos notice info nn Input list Input warn sqrt value_or append round time format info out fit time format std size mean info sum out predict _rank_layout isnan zeros sum range nan_to_num range join format name plot_model info percentage join format bias info len time format name _train_alternative_huang info format scipy_opt_minimize metrics _load_huang_context_with_matrix_and_weights_vector name getattr nan info float sum enumerate format name filter info zeros is_alternative metrics _check_for_non_finite_feature_vectors nan_to_num get_cached_mean_and_stdev nan AlternativeContextHuang zeros get_cached_value array enumerate dict sql_select defaultdict items get_one set _compute_prop_outer moderately_useful_helper BOXED GAUSSIAN get format _get_directory name PROPERTIES _compute_prop_inner info notice get_bad get list _sql_insert_curs _temporary_to_persistent_name _maybe_insert_pca _rename_files_batch append enum_from_json _roundabout get format sql_exec_curs propsdir str join enum_to_json makedirs relpath debug format rename sql_insert_curs round dump now flush load list format __name__ read debug append fail enumerate _find_tool items list _get_cache_directory _find_tool format _get_safe_command getenv warning info which split join quote map split append get_one _get_cache_directory getenv format warning info join list format info run randint append split dict environ devnull isnan max splitext list list le enumerate clear extend isnan isinstance isinf isinstance isinstance getattr isinstance join isspace all map partition append_common_all_constants send_tree_xml ElementTree Element Enum isinstance str sql_select get_one_or append_common_graph name classify fmtnum localized int sql_select_curs sql_select warning str Element append_common_all_constants send_tree_xml append_common_graph ElementTree _get_primary_axes_curs fmtnum sql_select_curs sql_select_curs _offer_picture_to_cache get_one_or end_headers _check_graphics_format str sql_select append send_response get send_header format _get_picture_from_cache _get_primary_axes flush write _get_conversion_command_from_svg call_graphstudy_tool graphstudy_manager call_image_magick OK len lower format _get_primary_axis_curs get_one_or sql_select_curs sqrt format token_hex debug rename dirname _get_cache_file_name makedirs debug format _get_cache_file_name items list Element _get_layout_ids_from_query _load_feature_rows rfc5322 set query parse_qs send_tree_xml graphstudy_manager attribute_projection ElementTree list query parse_qs send_tree_xml ElementTree set list all map get items list dict parse_bool startswith enum_from_json sorted _get_perf_stats_cooked items list min _get_perf_stats_raw mean append median sum max len append sql_select defaultdict frozenset query parse_qs _check_graphics_format get_bool_from_query sql_select format name _serve_plot query FIXED_COUNT parse_qs _get_bincount_from_query find_tool_lazily get_first_or _check_graphics_format list get_bool_from_query format find_tool_lazily _get_sigma_from_query _serve_plot query parse_qs median keys query parse_qs values _check_graphics_format list get_bool_from_query sorted name _serve_plot append format frozenset set lower items join dict filter graphstudy_manager send_header str flush write call_gnuplot encode send_response OK end_headers len int float get list all map lstrip list get_bool_from_query get_int_from_query tuple map sql_exec join str isinstance IntEnum name fmtnum upper append_child next isinstance IntEnum items list replace SubElement set join addprevious ProcessingInstruction int strip Id format runs stdout writeout cmd cross_validate print defaultdict format replace make_statistic dict keys len list stdev dict mean keys mean stdev dict exclude set include print_punctures print name sorted time message print_message store_info file getenv start Info update_progress_report load_info FIELDS join format fmttabs fmtmsg fmttdiff out _warning float print report format getenv report abspath expanduser getenv isabs which print format append join report quote prettytime sorted format all PASSED SKIPPED map FAILED report keys values stdin write_texdef print format_integer function invoke_argument_parser guess_archive_format cache serve_from_download abspath serve_from_cache urls extract format isfile output extract_files copyfile filter info upper format hexhash info format copyfile dirname info link enumerate makedirs format info callback read write getenv columns str lower fromhex partition strip partition list map print Tests rename values defaultdict load_rename_table name repr enum_to_json append pattern sum glob analysis_dict2tuple enum_from_json write_output bemoan output check_lazy_okay join sorted basename format_rfc5322 replace format items print stdev mean sqrt filter strip map split bemoan format getenv print basename basename format_rfc5322 test update sorted print_names keys dmnames list Id Tests map enum_from_json float enumerate split enum_from_json Tests partition replace corpus get_layer_dim_def fmtint round compile get_puncture_properties print_output escape add properties directory join sorted basename format_rfc5322 replace format items print set get_princomp input minor major run sqrt get log2 nan find_dependencies find_executable str columns tex get_tool_flags bib prep_deps find_tools deps idx prepare_dependencies run_toolchain find_deps isdir format quote getenv info which join format info split list format quote debug find_dependency dict append join isdir endswith isfile sep join items rstrip format quote relpath symlink dirname info sep makedirs format format_rfc5322 debug run_command splitext info range join str format time quote debug map returncode dict floor info environ run update sorted format quote md5 debug glob encode join sorted format quote glob escape info formatdate startswith split

<!-- -*- coding:utf-8; mode:markdown; -*- --> <!-- Copyright (C) 2018 Moritz Klammler <moritz.klammler@alumni.kit.edu> --> <!-- --> <!-- Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free --> <!-- Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with no --> <!-- Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. --> <!-- --> <!-- A copy of the license is included in the section entitled "GNU Free Documentation License". --> # Aesthetic Discrimination of Graph Layouts This package contains the all code for [Moritz Klammler's master's thesis](http://klammler.eu/msc/)

7-B/yoco

['style transfer']

['Deep Photo Style Transfer']

server.py app.py wsgi.py segmentation/torch_neural_style_transfer.py segmentation/merge_image.py convert.py segmentation/evaluate.py main add_header coloring github sketch image_resize convert_to_line png2svg simplify sobel main get_palette get_arguments image_resize image_loader StyleLoss Normalization gram_matrix imshow main ContentLoss get_style_model_and_losses run_style_transfer get_input_optimizer get join convert_to_line save filename join float resize CV_8U Sobel addWeighted image_resize GaussianBlur sobel imread system fromarray data time evaluate model print load_lua gpu_usage mean unsqueeze png2svg save_image forward std sketch join simplify mkdir add_argument ArgumentParser range load SCHPDataset Compose output DataParallel restore_weight eval load_state_dict DataLoader get_palette cuda get_arguments makedirs ceil unsqueeze open squeeze pause clone unloader title t size mm view deepcopy children format isinstance Sequential StyleLoss MaxPool2d add_module Conv2d len ReLU ContentLoss BatchNorm2d to range append detach LBFGS print clamp_ get_style_model_and_losses step get_input_optimizer ioff image_loader ToPILImage clone ion save_image resize to run_style_transfer open

<h1 align="center"> <br> <a href="https://141.223.140.22"><img src="img/YOCO-logo.png" alt="YOCO" width="200"><a> <br> </h1> ## Contents - [**Weekly Record**](https://github.com/7-B/yoco/wiki/Development-Record) - [**Reference**](https://github.com/7-B/yoco/wiki/%EC%B0%B8%EA%B3%A0-%EC%9E%90%EB%A3%8C) - [**Demo Video**](https://www.youtube.com/watch?v=Zw67sh-4jSI) <h4 align="center"> 사용자가 정한 이미지로 컬러링북을 제작합니다.</h4>

9ruddls3/CRNN_Pytorch

['optical character recognition', 'scene text recognition']

['An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition']

prepare.py train.py parms.py model/CRNN.py Preparing reset model_train BiDireRNN ToRNN CNN_block CRNN_model str list sort convert Compose tqdm abspath device append to listdir enumerate len hasattr isinstance xavier_normal fill_ Conv2d modules reset_parameters zero_ BatchNorm2d weight Linear train_model join backward print Variable train zero_grad range type loss_function IntTensor append dataset step init_hidden len

# CRNN_Pytorch 이 CRNN 패키지는 제 데이터셋을 기반으로 진행하였고, 구글 코랩을 기반으로 진행하였습니다. 전반적인 Hyper-Parameter는 prams.py 에서 선언되었으며, 이를 수정함으로써 튜닝 할 수 있습니다. 문자열 이미지데이터가 있고, 파일명이 'label.확장자'로 되어있다면, prepare.py 를 통해 Dataloader 로 변환 할 수 있습니다. This package is for my custom Dataset and based on Google Colaborative. If you use this for your custom dataset, you can convert your image folder to torchvision.Dataloader by using prepare.py * * * ## What is CRNN? Convolution layer를 이용해 이미지의 속성을 추출한 후, 추출한 이미지를 시계열 데이터로 간주하여 예측값을 predict 하는 기법 더 명확한 개념 설명은 아래 링크에서 확인 가능

A-Jacobson/Depth_in_The_Wild

['depth estimation']

['Single-Image Depth Perception in the Wild']

train.py datasets.py models.py train_utils.py criterion.py RelativeDepthLoss NYUDepth HourGlass ConvReluBN InceptionS InceptionL main validate AverageMeter save_checkpoint _fit_epoch prep_img fit NYUDepth load show state_dict plot xlabel ylabel RMSprop parameters save_checkpoint load_state_dict RelativeDepthLoss legend HourGlass cuda fit update format criterion model Variable backward AverageMeter zero_grad set_description avg tqdm_notebook train step cuda list format print DataLoader tqdm_notebook _fit_epoch range len update criterion model Variable AverageMeter eval DataLoader cuda dict copyfile save

# Depth in The Wild Pytorch implementation of Single-Image Depth Perception in the Wild https://arxiv.org/pdf/1604.03901.pdf - [x] Data Loader for NYUDepth - [x] Architecture - [x] Custom Criterion - [x] Train on Small sample - [x] Tweak Architecture - [ ] Fully Train Model (310/~500 epochs) - [ ] Validate Results ## Network Architecture:

A-ZHANG1/PSENet

['optical character recognition', 'scene text detection', 'curved text detection']

['Shape Robust Text Detection with Progressive Scale Expansion Network', 'Shape Robust Text Detection with Progressive Scale Expansion Network']

util/statistic.py util/tf.py pypse.py util/event.py train_ic15.py models/pvanet.py metrics.py util/feature.py util/proc.py test_ctw1500.py util/rand.py util/t.py util/neighbour.py util/caffe_.py models/__init__.py util/dtype.py dataset/icdar2015_loader.py util/ml.py util/str_.py utils.py util/test.py util/log.py test_ic15.py pse/.ycm_extra_conf.py pse/__init__.py util/misc.py eval/ic15/rrc_evaluation_funcs_v2.py pse/__main__.py util/__init__.py util/logger.py dataset/ctw1500_test_loader.py models/pvanet_1.py util/mask.py eval/ic15/rrc_evaluation_funcs_v1.py util/url.py dataset/ctw1500_loader.py eval/ctw1500/eval_ctw1500.py eval/ctw1500/file_util.py util/thread_.py models/fpn_resnet.py util/np.py util/dec.py util/io_.py util/mod.py train_ctw1500.py dataset/icdar2015_test_loader.py eval/ic15/rrc_evaluation_funcs.py eval/ic15/script.py util/cmd.py util/img.py dataset/__init__.py util/plt.py eval/ic15/file_util.py runningScore pse debug test write_result_as_txt extend_3c polygon_from_points debug test write_result_as_txt extend_3c polygon_from_points cal_kernel_score dice_loss ohem_batch adjust_learning_rate save_checkpoint ohem_single main cal_text_score train cal_kernel_score dice_loss ohem_batch adjust_learning_rate save_checkpoint ohem_single main cal_text_score train MultiCropEnsemble initvars get_img random_crop shrink get_bboxes random_rotate random_scale dist random_horizontal_flip scale CTW1500Loader perimeter get_img CTW1500TestLoader scale IC15Loader get_img random_crop shrink get_bboxes random_rotate random_scale dist random_horizontal_flip scale perimeter get_img IC15TestLoader scale get_union get_gt get_pred get_intersection write_file_not_cover write_file read_file read_dir write_file_not_cover write_file read_file read_dir validate_point_inside_bounds load_zip_file_keys validate_clockwise_points validate_lines_in_file decode_utf8 print_help main_validation get_tl_line_values load_zip_file get_tl_line_values_from_file_contents validate_tl_line main_evaluation validate_point_inside_bounds load_zip_file_keys validate_clockwise_points validate_lines_in_file decode_utf8 print_help main_validation get_tl_line_values load_zip_file get_tl_line_values_from_file_contents validate_tl_line main_evaluation validate_point_inside_bounds load_zip_file_keys validate_clockwise_points validate_lines_in_file decode_utf8 print_help main_validation get_tl_line_values load_zip_file get_tl_line_values_from_file_contents validate_tl_line main_evaluation evaluation_imports evaluate_method default_evaluation_params validate_data ResNet resnet50 Bottleneck resnet152 conv3x3 resnet34 resnet18 BasicBlock resnet101 ConvBnAct pvanet mCReLU_residual Inception PVANetFeat CReLU mCReLU_base PVANet ConvBnAct pvanet mCReLU_residual Inception PVANetFeat CReLU mCReLU_base PVANet GetCompilationInfoForFile IsHeaderFile MakeRelativePathsInFlagsAbsolute FlagsForFile DirectoryOfThisScript pse get_data get_params draw_log cmd print_calling_in_short_for_tf timeit print_calling print_test print_calling_in_short is_tuple int is_number is_str cast is_list double wait_key hog get_contour_min_area_box blur imwrite get_rect_iou black get_value put_text bgr2rgb get_roi render_points bgr2gray get_contour_region_iou resize convex_hull draw_contours get_contour_rect_box get_shape set_value is_in_contour is_valid_jpg move_win get_contour_region_in_rect fill_bbox imshow apply_mask random_color_3 imread bilateral_blur find_contours points_to_contours maximize_win rect_area rgb2bgr contour_to_points ds_size points_to_contour eq_color find_two_level_contours get_wh average_blur rgb2gray get_contour_region_in_min_area_rect rotate_point_by_90 white filter2D is_white translate get_contour_area rectangle min_area_rect rect_perimeter get_rect_points rotate_about_center gaussian_blur circle get_dir search is_dir dump_mat read_h5_attrs exists get_filename cd join_path load_mat get_file_size get_absolute_path cat dir_mat dump_json create_h5 dump pwd copy is_path mkdir ls open_h5 load remove write_lines read_h5 make_parent_dir find_files read_lines get_date_str init_logger plot_overlap savefig Logger LoggerMonitor find_black_components find_white_components init_params AverageMeter mkdir_p get_mean_and_std kmeans try_import_by_name add_ancester_dir_to_path import_by_name is_main get_mod_by_name add_to_path load_mod_from_path n2 _in_image count_neighbours get_neighbours n1 n1_count n8 n2_count n4 norm2_squared smooth flatten empty_list norm2 sum_all angle_with_x has_infty sin arcsin norm1 eu_dist iterable is_2D shuffle cos_dist chi_squared_dist clone is_empty has_nan has_nan_or_infty show plot_solver_data line show_images get_random_line_style to_ROI draw imshow rectangle hist maximize_figure set_subtitle save_image get_pid kill wait_for_pool get_pool cpu_count set_proc_name ps_aux_grep shuffle normal randint sample D E join index_of find_all is_none_or_empty ends_with remove_all remove_invisible to_lowercase starts_with is_str int_array_to_str contains to_uppercase split replace_all add_noise crop_into get_latest_ckpt get_init_fn gpu_config Print is_gpu_available min_area_rect focal_loss_layer_initializer get_variable_names_in_checkpoint sum_gradients get_all_ckpts get_update_op get_variables_to_train focal_loss get_iter get_available_gpus wait_for_checkpoint get_current_thread ThreadPool create_and_start ProcessPool is_alive get_current_thread_name download argv cit get_count exit sit connectedComponents get transpose copy put shape Queue zeros range len reshape concatenate imwrite concatenate print makedirs append range len join_path makedirs write_lines append range enumerate len int T empty model pse sign DataLoader resize resnet152 cuda max RETR_TREE list OrderedDict shape load_state_dict resnet101 append range format min_kernel_area resnet50 synchronize findContours astype debug copy write_result_as_txt mean eval img_paths resume scale flush enumerate load items time uint8 drawContours binary_th CHAIN_APPROX_SIMPLE print Variable reshape float32 sigmoid parameters CTW1500TestLoader isfile zeros tuple pypse minAreaRect boxPoints IC15TestLoader numpy int sort min astype sum concatenate ohem_single append float numpy range sigmoid sum view mean update astype int32 get_scores numpy update astype int32 get_scores numpy model zero_grad runningScore cuda cal_kernel_score step append cal_text_score sum range update format size astype ohem_batch item float flush enumerate time criterion backward Variable print AverageMeter numpy len param_groups lr join save SGD DataLoader adjust_learning_rate save_checkpoint Logger resnet152 cuda hasattr load_state_dict resnet101 append module range pretrain resnet50 close lr resume CTW1500Loader checkpoint optimizer flush join load print parameters n_epoch train set_names makedirs summary_path IC15Loader SummaryWriter pvanet add_scalar fill_ isinstance out_channels Conv2d normal_ sqrt zero_ BatchNorm2d imread int asarray remove_all append read_lines split range copy len warpAffine random getRotationMatrix2D range len max resize max min choice resize array min where randint max range len range PyclipperOffset int JT_ROUND area min append AddPath array perimeter ET_CLOSEDPOLYGON Execute print append split append int asarray split area append sort walk replace read close open join makedirs close write open join makedirs close write open write exit group match namelist append ZipFile group match namelist append ZipFile decode BOM_UTF8 replace startswith encode validate_tl_line decode_utf8 replace split get_tl_line_values validate_point_inside_bounds int replace group match replace argsort append get_tl_line_values split update default_evaluation_params_fn validate_data_fn writestr list items write dumps close dict print_help evaluate_method_fn ZipFile makedirs update default_evaluation_params_fn validate_data_fn print exit dict validate_clockwise_points float load_zip_file validate_lines_in_file compute_ap area list decode_utf8 append polygon_from_points range import_module get_intersection_over_union load_zip_file empty get_pred get_tl_line_values_from_file_contents float items namedtuple int8 rectangle_to_polygon Rectangle get_intersection zeros len load_url ResNet load_state_dict load_url ResNet load_state_dict list ResNet load_url load_state_dict keys state_dict list ResNet load_url load_state_dict keys state_dict list ResNet load_url load_state_dict keys state_dict PVANet append join startswith IsHeaderFile compiler_flags_ exists compiler_flags_ GetCompilationInfoForFile compiler_working_dir_ MakeRelativePathsInFlagsAbsolute DirectoryOfThisScript cpse array net Solver isinstance append net Solver isinstance show int get_random_line_style plot print readlines smooth len contains eval get_absolute_path save_image plt legend append float open isinstance debug waitKey ord bgr2rgb get_absolute_path wait_key namedWindow isinstance destroyAllWindows move_win rgb2bgr WINDOW_NORMAL imread maximize_win get_absolute_path rgb2bgr make_parent_dir moveWindow setWindowProperty WND_PROP_FULLSCREEN enumerate get_shape get_shape min max drawContours boundingRect get_contour_rect_box minAreaRect BoxPoints int0 get_shape get_contour_rect_box warpAffine int BoxPoints transpose hstack getRotationMatrix2D dot get_roi minAreaRect points_to_contour black draw_contours shape to_contours draw_contours assert_equal asarray range GaussianBlur bilateralFilter putText int32 FONT_HERSHEY_SIMPLEX get_shape int tuple warpAffine get_wh float32 cos deg2rad getRotationMatrix2D dot sin abs array _get_area transpose _get_inter zeros range len findContours asarray copy findContours copy pointPolygonTest convexHull randint list asarray range zip minAreaRect empty points_to_contour get_absolute_path makedirs get_dir mkdir get_absolute_path get_dir mkdir get_absolute_path get_absolute_path get_absolute_path is_dir get_absolute_path expanduser startswith chdir get_absolute_path append listdir get_absolute_path ends_with get_absolute_path open get_absolute_path make_parent_dir get_absolute_path get_absolute_path get_absolute_path savemat make_parent_dir get_absolute_path getsize get_absolute_path get_absolute_path make_parent_dir get_absolute_path get_absolute_path get_absolute_path join_path extend ls is_dir get_absolute_path find_files append get_absolute_path make_parent_dir now setFormatter basicConfig print join_path addHandler make_parent_dir StreamHandler get_date_str Formatter setLevel asarray arange plot numbers enumerate len pop black set_root insert copy get_neighbours N4 shape get_new_root get_root append set_visited range is_visited print DataLoader div_ zeros range len normal constant isinstance kaiming_normal Conv2d bias modules BatchNorm2d weight Linear makedirs asarray warn flatten append enumerate insert join_path add_to_path get_dir __import__ import_by_name get_absolute_path get_filename append _in_image append _in_image append _in_image append _in_image norm2 zip shape asarray extend len pi asarray asarray reshape flatten sqrt shape range shape asarray has_infty has_nan enumerate len show asarray join_path flatten linspace figure save_image load show val_accuracies list plot training_losses val_losses training_accuracies figure legend range len Rectangle add_patch full_screen_toggle get_current_fig_manager add_line Line2D linspace show_images show set_title set_subtitle axis colorbar bgr2rgb imshow maximize_figure subplot2grid append save_image enumerate get_absolute_path savefig imsave make_parent_dir set_xlim set_ylim suptitle maximize_figure randint len Pool join close setproctitle print get_pid cmd append int cmd split flatten flatten append pop list tuple to_lowercase is_str enumerate list tuple to_lowercase is_str enumerate to_lowercase findall replace replace_all binomial list_local_devices get_checkpoint_state is_dir get_absolute_path model_checkpoint_path get_checkpoint_state all_model_checkpoint_paths int get_latest_ckpt is_none_or_empty latest_checkpoint print get_model_variables startswith info append extend get_collection TRAINABLE_VARIABLES get_latest_ckpt NewCheckpointReader get_variable_to_shape_map dtype set_shape py_func ConfigProto ones_like zeros_like sigmoid_cross_entropy_with_logits float32 where reduce_sum sigmoid pow cast stop_gradient name reduce_mean add_n histogram zip append scalar UPDATE_OPS get_collection start Thread setName print urlretrieve stat show_images get_count imwrite get_count imwrite asarray

# Shape Robust Text Detection with Progressive Scale Expansion Network ## Requirements * Python 2.7 * PyTorch v0.4.1+ * pyclipper * Polygon2 * OpenCV 3.4 (for c++ version pse) * opencv-python 3.4 ## Introduction Progressive Scale Expansion Network (PSENet) is a text detector which is able to well detect the arbitrary-shape text in natural scene.

AADeLucia/gpt2-narrative-decoding

['response generation']

['Decoding Methods for Neural Narrative Generation']

data/count_length.py mturk_human_evaluation/fleiss.py mturk_human_evaluation/cohen.py evaluation.py generate_responses.py mturk_human_evaluation/format_generated_narratives_csv.py data/preproc.py mturk_human_evaluation/create_html_survey.py sentBERT distinct_2 distinct_1 clean_response parse_args parse_args batchify print_kappas create_option_list create_evaluation_display create_examples_display create_narrative_display get_idxs fleiss_kappa format_for_html add_argument ArgumentParser encode pdist set len split list extend zip split debug strip lower sub findall append enumerate device print format items list sub list keys enumerate shape float sum multiply update split enumerate set strip sub

# Decoding Methods for Neural Narrative Generation Alexandra DeLucia\*, Aaron Mueller\*, Xiang "Lisa" Li, João Sedoc This repository contains code for replicating the approach and results of the paper [Decoding Strategies for Neural Narrative Generation](https://arxiv.org/abs/2010.07375). ## Models Our GPT-2 Medium models fine-tuned on the medium and large datasets are available on the library. If you would like the GPT-2 Small model, please reach out to Alexandra DeLucia. ``` model = AutoModel.from_pretrained("aadelucia/GPT2_medium_narrative_finetuned_medium") ``` ## Preprocessing Our preprocessing script may be found in `data/preproc.py`. To replicate our setup, download the [writingPrompts dataset](https://dl.fbaipublicfiles.com/fairseq/data/writingPrompts.tar.gz) and unzip the .tar into the `data` folder. Then run `preproc.py`.

AI-HPC-Research-Team/AIPerf

['automl']

['AIPerf: Automated machine learning as an AI-HPC benchmark']

examples/trials/network_morphism/cifar10/distributed_utils.py src/sdk/pynni/tests/models/pytorch_models/__init__.py tools/nni_cmd/command_utils.py tools/nni_cmd/url_utils.py tools/nni_annotation/testcase/usercode/nas.py src/nni_manager/training_service/test/mockedTrial.py src/sdk/pynni/tests/test_compressor.py src/sdk/pynni/setup.py src/sdk/pynni/tests/test_trial.py scripts/reports/save_log.py src/sdk/pynni/tests/test_protocol.py src/sdk/pynni/nni/networkmorphism_tuner/bayesian.py src/sdk/pynni/nni/networkmorphism_tuner/test_networkmorphism_tuner.py src/sdk/pynni/nni/nas_utils.py src/sdk/pynni/nni/hyperopt_tuner/hyperopt_tuner.py src/sdk/pynni/nni/trial.py src/sdk/pynni/tests/test_pruners.py src/sdk/pynni/nni/networkmorphism_tuner/networkmorphism_tuner.py src/sdk/pynni/nni/protocol.py examples/trials/network_morphism/imagenet/dataset.py src/sdk/pynni/nni/msg_dispatcher.py src/sdk/pynni/nni/networkmorphism_tuner/layer_transformer.py tools/nni_annotation/testcase/usercode/non_annotation/bar.py src/sdk/pynni/nni/utils.py setup.py tools/nni_cmd/updater.py tools/nni_cmd/config_schema.py examples/trials/network_morphism/cifar10/cifar10_keras_multigpu_ps.py tools/nni_gpu_tool/gpu_metrics_collector.py tools/nni_annotation/testcase/usercode/dir/simple.py src/sdk/pynni/nni/platform/local.py src/sdk/pynni/nni/common.py src/sdk/pynni/nni/tuner.py src/sdk/pynni/tests/__init__.py src/nni_manager/core/test/assessor.py src/sdk/pynni/nni/networkmorphism_tuner/layers.py tools/nni_annotation/testcase/annotated/dir/simple.py scripts/reports/score.py tools/nni_cmd/constants.py tools/nni_cmd/launcher.py src/sdk/pycli/setup.py tools/nni_annotation/testcase/usercode/non_annotation/foo.py scripts/reports/gen_report.py src/sdk/pynni/tests/models/pytorch_models/naive.py scripts/deploy/data_transmission/sshRemoteCmd.py tools/nni_cmd/launcher_utils.py examples/trials/network_morphism/cifar10/cifar10_pytorch.py src/sdk/pynni/nni/platform/test.py tools/nni_annotation/examples/mnist_generated.py tools/nni_cmd/nnictl_utils.py tools/nni_annotation/testcase/usercode/mnist.py scripts/build_data/preprocess_imagenet_validation_data.py src/sdk/pynni/tests/test_utils.py scripts/reports/report.py examples/trials/network_morphism/imagenet/imagenet_preprocessing.py src/sdk/pynni/tests/test_smartparam.py tools/nni_annotation/examples/mnist_without_annotation.py tools/nni_trial_tool/url_utils.py tools/nni_annotation/__init__.py tools/nni_trial_tool/constants.py examples/trials/network_morphism/imagenet/resource_monitor.py tools/nni_trial_tool/hdfsClientUtility.py examples/trials/network_morphism/imagenet/utils.py src/sdk/pynni/tests/test_assessor.py src/sdk/pynni/tests/test_msg_dispatcher.py tools/nni_trial_tool/test/test_hdfsClientUtility.py src/sdk/pynni/tests/models/pytorch_models/nested.py src/nni_manager/core/test/dummy_tuner.py src/sdk/pynni/nni/constants.py src/sdk/pynni/nni/networkmorphism_tuner/nn.py tools/nni_annotation/search_space_generator.py tools/nni_annotation/examples/mnist_with_annotation.py src/sdk/pynni/nni/platform/standalone.py src/sdk/pynni/nni/parameter_expressions.py src/sdk/pynni/tests/models/pytorch_models/mutable_scope.py src/sdk/pynni/nni/msg_dispatcher_base.py examples/trials/network_morphism/cifar10/cifar10_pytorch_allreduce.py src/sdk/pycli/nnicli/__init__.py src/nni_manager/core/test/dummy_assessor.py src/sdk/pynni/nni/__init__.py tools/nni_cmd/rest_utils.py src/sdk/pynni/nni/platform/__init__.py tools/setup.py src/sdk/pynni/nni/recoverable.py examples/trials/network_morphism/cifar10/resource_monitor.py scripts/build_data/build_imagenet_data.py tools/nni_cmd/nnictl.py tools/nni_trial_tool/log_utils.py tools/nni_cmd/ssh_utils.py tools/nni_cmd/common_utils.py tools/nni_annotation/testcase/annotated/non_annotation/bar.py examples/trials/network_morphism/imagenet/demo.py src/sdk/pynni/nni/assessor.py scripts/reports/profiler.py tools/nni_annotation/testcase/annotated/non_annotation/foo.py tools/nni_cmd/package_management.py tools/nni_cmd/tensorboard_utils.py tools/nni_annotation/testcase/annotated/handwrite.py src/sdk/pycli/nnicli/nni_client.py src/sdk/pynni/nni/networkmorphism_tuner/graph.py src/sdk/pynni/nni/networkmorphism_tuner/utils.py src/sdk/pynni/nni/__main__.py src/sdk/pynni/nni/smartparam.py tools/nni_trial_tool/rest_utils.py src/sdk/pynni/tests/test_builtin_tuners.py src/sdk/pynni/nni/networkmorphism_tuner/graph_transformer.py tools/nni_annotation/specific_code_generator.py tools/nni_annotation/testcase/annotated/mnist.py examples/trials/network_morphism/imagenet/imagenet_train.py tools/nni_cmd/config_utils.py src/sdk/pynni/nni/env_vars.py tools/nni_annotation/code_generator.py tools/nni_trial_tool/trial_keeper.py tools/nni_annotation/test_annotation.py src/sdk/pynni/nni/hyperopt_tuner/test_hyperopt_tuner.py src/sdk/pynni/tests/test_nas.py examples/trials/network_morphism/cifar10/utils.py tools/nni_annotation/testcase/annotated/nas.py read build_graph_from_json get_args train_eval parse_rev_args SendMetrics Epoch_num_record build_graph_from_json get_args parse_rev_args test train average_gradients broadcast_params DistModule dist_init record_device_info get_args test_record_device_info socket_client record_net_info write_file Cutout get_mean_and_std EarlyStopping data_transforms_mnist data_transforms_cifar10 init_params _parse_example_proto parse_record process_record_dataset get_filenames parse_rev_args build_graph_from_json get_args train_eval _aspect_preserving_resize _decode_crop_and_flip _central_crop _smallest_size_at_least _mean_image_subtraction _resize_image preprocess_image build_graph_from_json get_args train_eval parse_rev_args SendMetrics record_device_info get_args test_record_device_info record_net_info write_file MinGpuMem EarlyStopping predict_acc trial_activity ImageCoder _convert_to_example _process_image_files _process_image _int64_feature _process_dataset _is_cmyk _find_image_files _build_bounding_box_lookup _find_human_readable_labels _build_synset_lookup _find_image_bounding_boxes _bytes_feature _float_feature main _process_image_files_batch _is_png ssh_cmd read_file get_args GenPerfdata GenReport cal_add_flops cal_conv_flops cal_trial_flops_per_image cal_bn_flops cal_relu_flops cal_avgpool_flops profiler cal_dense_flops cal_maxpool_flops cal_softmax_flops main get_args formatnum main_grid display_log save_log cal_report_results find_max_acc find_all_trials find_startime conversion_time process_log receive send DummyAssessor DummyTuner user_code sdk_send_data get_trial_job list_trial_jobs export_data get_job_metrics get_experiment_profile get_experiment_status stop_nni set_endpoint _check_endpoint start_nni _http_succeed version get_job_statistics _nni_rest_get _create_process read Assessor AssessResult enable_multi_phase multi_thread_enabled enable_multi_thread init_standalone_logger init_logger multi_phase_enabled _LoggerFileWrapper _load_env_vars _create_parameter_id _pack_parameter _sort_history MsgDispatcher MsgDispatcherBase enas_mode darts_mode reload_tensorflow_variables training_update _decompose_general_key _get_layer_and_inputs_from_tuner oneshot_mode convert_nas_search_space rewrite_nas_space darts_training classic_mode _construct_general_key qloguniform normal choice qlognormal uniform qnormal lognormal randint loguniform quniform CommandType receive send Recoverable qloguniform normal mutable_layer choice qlognormal uniform qnormal lognormal function_choice randint loguniform quniform report_intermediate_result get_sequence_id get_experiment_id get_next_parameter report_final_result get_current_parameter get_trial_id Tuner json2space split_index OptimizeMode extract_scalar_reward MetricType json2parameter extract_scalar_history init_dispatcher_logger NodeType convert_dict2tuple NoMoreTrialError _create_tuner augment_classargs _create_assessor create_customized_class_instance _run_advisor create_builtin_class_instance main json2space json2parameter _add_index HyperoptTuner json2vals HyperoptTunerTestCase skip_connections_distance ReverseElem layers_distance vector_distance job SearchTree skip_connection_distance Elem layer_distance edit_distance bourgain_embedding_matrix contain_old attribute_difference BayesianOptimizer IncrementalGaussianProcess edit_distance_matrix contain NetworkDescriptor graph_to_onnx json_to_graph JSONModel Graph TorchModel TfModel ONNXModel onnx_to_graph Node KerasModel graph_to_json to_skip_connection_graph to_deeper_graph to_wider_graph to_skip_connection_graph2 create_new_layer legal_graph to_deeper_graph2 transform StubFlatten is_layer get_conv_class get_global_avg_pooling_class StubLayer get_dropout_class set_stub_weight_to_torch StubConv2d StubGlobalPooling1d StubConv3d StubBatchNormalization1d StubBatchNormalization2d get_batch_norm_class TorchConcatenate StubWeightBiasLayer get_pooling_class GlobalAvgPool1d set_keras_weight_to_stub StubBatchNormalization3d StubGlobalPooling set_stub_weight_to_keras StubConv StubDense StubDropout2d tf_dropout layer_width set_torch_weight_to_stub StubPooling3d StubConcatenate StubDropout StubGlobalPooling2d StubDropout1d StubDropout3d TorchAdd StubBatchNormalization TorchFlatten GlobalAvgPool2d StubAdd StubSoftmax StubPooling1d keras_dropout to_real_tf_layer StubReLU StubInput layer_description_builder layer_description_extractor StubGlobalPooling3d StubPooling2d get_n_dim StubAggregateLayer AvgPool to_real_keras_layer StubPooling GlobalAvgPool3d StubConv1d wider_pre_conv add_noise deeper_conv_block dense_to_deeper_block init_bn_weight wider_next_conv init_dense_weight wider_pre_dense wider_bn init_conv_weight wider_next_dense NetworkMorphismTuner NetworkGenerator MlpGenerator CnnGenerator NetworkMorphismTestCase Constant send_metric get_sequence_id get_experiment_id get_next_parameter request_next_parameter get_trial_id send_metric get_sequence_id get_experiment_id get_next_parameter get_trial_id send_metric get_sequence_id get_experiment_id get_next_parameter get_last_metric get_trial_id init_params AssessorTestCase _restore_io NaiveAssessor _reverse_io BuiltinTunersTestCase get_tf_model TorchModel CompressorTestCase tf2 _restore_io MsgDispatcherTestCase _reverse_io NaiveTuner NasTestCase ProtocolTestCase _prepare_receive _prepare_send PrunerTestCase Model validate_sparsity pruners_test foo bar SmartParamTestCase conv2D TrialTestCase UtilsTestCase Node Cell SpaceWithMutableScope Layer NaiveSearchSpace NestedSpace MutableOp Transformer test_variable_equal parse FuncReplacer parse_annotation replace_variable_node parse_annotation_function parse_nni_variable convert_args_to_dict parse_nni_function parse_annotation_mutable_layers replace_function_node make_lambda SearchSpaceGenerator generate Transformer test_variable_equal parse FuncReplacer parse_annotation replace_variable_node parse_annotation_function parse_nni_variable convert_args_to_dict parse_nni_function parse_annotation_mutable_layers replace_function_node make_lambda AnnotationTestCase generate_search_space _generate_file_search_space expand_annotations _generate_specific_file _expand_file_annotations bias_variable max_pool download_mnist_retry conv2d generate_defualt_params main avg_pool MnistNetwork weight_variable bias_variable max_pool download_mnist_retry conv2d generate_defualt_params main avg_pool MnistNetwork weight_variable bias_variable max_pool download_mnist_retry conv2d generate_defualt_params main avg_pool MnistNetwork weight_variable max_pool bias_variable generate_default_params max_pool conv2d main avg_pool MnistNetwork weight_variable add_one add_three add_four main add_two max_pool bar bias_variable generate_default_params max_pool conv2d main avg_pool MnistNetwork weight_variable add_one add_three add_four main add_two max_pool bar check_output_command install_requirements_command _get_pip_install kill_command install_package_command get_nni_installation_path print_normal check_tensorboard_version print_error get_python_dir get_json_content get_yml_content detect_port print_warning detect_process get_user setChoice setType setNumberRange setPathCheck Config Experiments set_platform_config set_remote_config set_frameworkcontroller_config set_dlts_config set_pai_config manage_stopped_experiment set_kubeflow_config get_log_path start_rest_server set_pai_yarn_config launch_experiment print_log_content start_monitor resume_experiment setNNIManagerIp create_experiment set_trial_config set_local_config set_experiment view_experiment validate_search_space_content validate_machine_list parse_tuner_content validate_kubeflow_operators parse_path validate_common_content parse_advisor_content parse_assessor_content parse_time validate_pai_config_path validate_annotation_content validate_customized_file validate_pai_trial_conifg parse_relative_path validate_all_content expand_path parse_args nni_info update_experiment convert_time_stamp_to_date trial_ls search_space_auto_gen hdfs_clean trial_kill log_trial stop_experiment experiment_list webui_url get_experiment_time remote_clean local_clean get_experiment_status get_config_filename experiment_clean list_experiment get_time_interval get_experiment_port get_config webui_nas export_trials_data check_rest check_experiment_id log_internal experiment_status trial_codegen parse_ids show_experiment_info log_stdout monitor_experiment platform_clean get_platform_dir set_monitor log_stderr package_install package_show process_install check_rest_server rest_get rest_post rest_put check_rest_server_quick rest_delete check_response remove_remote_directory create_ssh_sftp_client copy_remote_directory_to_local check_environment format_tensorboard_log_path start_tensorboard get_path_list parse_log_path copy_data_from_remote start_tensorboard_process stop_tensorboard import_data_to_restful_server load_search_space update_concurrency import_data get_query_type validate_file validate_dispatcher update_searchspace validate_digit update_duration update_trialnum update_experiment_profile trial_job_id_url import_data_url experiment_url cluster_metadata_url export_data_url tensorboard_url check_status_url get_local_urls trial_jobs_url parse_nvidia_smi_result main check_ready_to_run gen_empty_gpu_metric copyHdfsFileToLocal copyFileToHdfs copyHdfsDirectoryToLocal copyDirectoryToHdfs StdOutputType nni_log RemoteLogger LogType PipeLogReader NNIRestLogHanlder rest_post rest_put rest_delete rest_get download_parameter fetch_parameter_file check_version get_hdfs_client is_multi_phase main_loop trial_keeper_help_info gen_send_version_url gen_parameter_meta_url gen_send_stdout_url HDFSClientUtilityTest add_argument ArgumentParser debug produce_keras_model json_to_graph operation_history build_graph_from_json str multi_gpu_model debug to_categorical astype write close load_data split open len SGD open str Adam RMSprop append Adagrad debug Adamax close get_trial_id compile int Adadelta evaluate fit write split len produce_torch_model Adagrad Adadelta Adamax Adam SGD RMSprop parameters DataLoader data_transforms_cifar10 DataParallel CIFAR10 to CrossEntropyLoss criterion backward debug zero_grad step max net enumerate debug eval all_reduce parameters requires_grad data list values broadcast int str init_process_group set_device write close get_world_size set_start_method device_count get_rank open sleep str get_recv bytes_recv get_send write close sleep bytes_sent open close write open str join cpu_percent strip readlines sleep write_file used virtual_memory print sendto decode sleep Cutout cutout Compose cutout_length append Cutout cutout Compose cutout_length append print DataLoader div_ zeros range len normal constant isinstance kaiming_normal Conv2d bias modules BatchNorm2d weight Linear append join listdir str map_and_batch shuffle apply repeat prefetch parse_single_example one_hot cast int32 _parse_example_proto preprocess_image cast load close dumps produce_tf_model open MirroredStrategy batch_size TFRecordDataset warmup_1 rand train_data_dir warmup_3 parse_rev_args flat_map from_tensor_slices val_data_dir LearningRateScheduler get_filenames warmup_2 process_record_dataset min slave epochs sample_distorted_bounding_box constant random_flip_left_right decode_and_crop_jpeg extract_jpeg_shape stack unstack shape expand_dims cast float32 minimum int32 cast float32 shape _smallest_size_at_least _aspect_preserving_resize _decode_crop_and_flip multiply cast _central_crop _resize_image decode_jpeg int list get_experiment_id range float makedirs predict_acc ModelCheckpoint int enumerate str ppid_gpu_info system get_md5 sleep ppid_cpu_info max list str print subtract square curve_fit sqrt append sum array logfun len list readlines round min Example read print _is_cmyk cmyk_to_rgb png_to_jpeg _is_png decode_jpeg int join _convert_to_example arange _process_image TFRecordWriter print astype write SerializeToString close output_directory range flush len int ImageCoder Thread join print astype Coordinator start append range flush len seed list print Glob extend shuffle range len append append basename print _find_image_files _process_image_files _find_human_readable_labels labels_file readlines split print readlines append float split _process_dataset imagenet_metadata_file print train_directory validation_directory _build_synset_lookup train_shards output_directory validation_shards print exit str AutoAddPolicy connect close set_missing_host_key_policy SSHClient exec_command load int cal_conv_flops cal_bn_flops cal_add_flops cal_relu_flops close cal_avgpool_flops len cal_dense_flops cal_maxpool_flops cal_softmax_flops range open append join cal_trial_flops_per_image listdir subplot join plot xlabel grid close ylabel set_visible savefig figure set_major_formatter tick_params array cal_report_results display_log id main_grid join system mkdir range len join readlines close extend len append range open fromtimestamp join mktime strptime timestamp timedelta fromtimestamp min timestamp timedelta isfile range len int format arange find_max_acc e find_startime profiler log append float abs range len join list sorted isfile find_all_trials min dict process_log zip listdir keys range values len encode write flush int read decode getenv join len format range sdk_send_data sleep get format status_code _check_endpoint _http_succeed print readline strip Popen split split get setFormatter getLogger addHandler StreamHandler localtime Formatter _LoggerFileWrapper setLevel INFO open stdout setFormatter basicConfig getLogger addHandler StreamHandler Formatter setLevel INFO append enumerate acquire release list keys _get_layer_and_inputs_from_tuner get_next_parameter list format case len boolean_mask dict range get_variable get_shape list dropout get_next_parameter reduce_sum values len list format softmax add_n append values get_variable load sorted list format _get_layer_and_inputs_from_tuner _decompose_general_key get_next_parameter add set index warning split MomentumOptimizer compute_gradients apply_gradients run darts_training reload_tensorflow_variables get_current_parameter info get update list items isinstance error len dict warning info _construct_general_key debug acquire debug CommandType Process time int recv_pyobj strip start history acquire popen send_pyobj release to_json send_metric send_pyobj recv_pyobj isinstance isinstance isinstance join NNI_LOG_DIRECTORY NNI_LOG_LEVEL init_logger list isinstance append keys enumerate deepcopy list str isinstance len dict append randint keys enumerate items list class_constructor import_module getattr augment_classargs get class_constructor import_module getattr append get decode _create_tuner enable_multi_phase debug add_argument dumps enable_multi_thread parse_known_args _create_assessor MsgDispatcher loads ArgumentParser _run_advisor _on_exit run create_customized_class_instance get create_builtin_class_instance run create_customized_class_instance get create_builtin_class_instance create_customized_class_instance get create_builtin_class_instance deepcopy str choice dict randint log clip str list isinstance keys enumerate items list isinstance dict enumerate is_layer min layer_distance zeros range len abs zeros skip_connection_distance enumerate layers_distance layers edit_distance zeros enumerate array seed int list min log choice ceil array range append len join partial close map Pool range len produce_onnx_model save Graph parsing_onnx_model produce_json_model dumps items list int Graph tuple layer_description_builder dict Node loads append list is_layer to_wider_model units wide_layer_ids sample filters sorted sample skip_connection_layer_ids to_add_skip_model append to_concat_skip_model range len is_layer sample skip_connection_layer_ids to_add_skip_model append to_concat_skip_model range len layer_class is_layer shape DENSE_DROPOUT_RATE StubDense n_dim create_new_layer to_deeper_model deep_layer_ids sample layer_class get_conv_class StubReLU n_dim to_deeper_model shape deep_layer_ids2 get_layers_id sample get_batch_norm_class skip_connections extract_descriptor deepcopy N_NEIGHBOURS to_skip_connection_graph2 to_wider_graph to_deeper_graph2 randrange append range shape len is_layer shape len is_layer input list output isinstance startswith list isinstance is_layer import_weights import_weights_keras export_weights export_weights_keras isinstance tuple get_n_dim zeros filters range set_weights units eye zeros StubDense set_weights add_noise concatenate units copy input_units append randint get_weights StubDense range set_weights concatenate set_weights kernel_size filters copy get_n_dim stride append randint get_weights input_channel enumerate list concatenate tuple filters get_n_dim shape zeros get_weights input_channel set_weights num_features concatenate tuple copy get_n_dim zip get_weights set_weights int concatenate units copy input_units zeros get_weights StubDense set_weights ptp flatten uniform shape units zeros eye set_weights tuple get_n_dim zeros filters range set_weights num_features set_weights to_json send_metric load join request_next_parameter sleep open name print write close run encode flush open warning error loads info deepcopy loads seek seek Sequential compile BytesIO BytesIO format bias_mask numel item append remove format compress randn print model backward zero_grad export_model SGD parameters Model v step long cross_entropy strip elts Attribute values str Name Num append Dict parse value Str List Assign Call keywords func zip keys args parse parse_annotation value strip parse_annotation_function convert_args_to_dict append keyword convert_args_to_dict strip Str parse_annotation_function str Str strip append Dict args arguments AST list items isinstance parse_nni_variable visit parse_nni_function Transformer Import insert visit body enumerate visit parse SearchSpaceGenerator index update join replace endswith _generate_file_search_space walk len join replace endswith makedirs copyfile walk len truncated_normal constant range add_graph mkdtemp FileWriter download_mnist_retry get_default_graph build_network MnistNetwork read_data_sets mutable_layer report_intermediate_result report_final_result sleep format exit exists print_error call Process CTRL_BREAK_EVENT send_signal call _get_pip_install call _get_pip_install append print print print Process socket close connect AF_INET SOCK_STREAM join getusersitepackages try_installation_path_sequentially print_error exit getenv isfile join print check_output_command print_normal get_log_path strip system popen get_nni_installation_path join print_normal print_error exit detect_port get_log_path format print text rest_put dumps get_log_path dict cluster_metadata_url check_response get str text rest_put dumps dict cluster_metadata_url get_log_path get str isinstance text rest_put dumps get_log_path dict setNNIManagerIp cluster_metadata_url range len text rest_put dumps dict cluster_metadata_url get_log_path text rest_put dumps dict setNNIManagerIp cluster_metadata_url get_log_path text rest_put dumps dict setNNIManagerIp cluster_metadata_url get_log_path text rest_put dumps dict setNNIManagerIp cluster_metadata_url get_log_path text rest_put dumps dict setNNIManagerIp cluster_metadata_url get_log_path text rest_put dumps dict setNNIManagerIp cluster_metadata_url get_log_path get str format print_error rest_post text dumps get_log_path dict experiment_url append check_response pid print_normal set_frameworkcontroller_config format set_pai_yarn_config set_remote_config set_local_config print_error exit kill_command set_dlts_config set_pai_config set_kubeflow_config Config decode pid set_platform_config print_error port get_local_urls get_log_path start_rest_server add_experiment check_rest_server generate_search_space print_normal print_log_content mkdtemp exit start_monitor foreground get debug expand_annotations gettempdir get_json_content kill_command set_config join Experiments print dumps set_experiment get_user makedirs Config join config launch_experiment print_error exit get_yml_content port abspath sample validate_all_content ascii_letters digits set_config update_experiment Experiments Config format print_normal join launch_experiment print_error get_config exit id port get_all_experiments sample ascii_letters digits set_config manage_stopped_experiment manage_stopped_experiment get expanduser get join print_normal exit print_error get dirname parse_relative_path expand_path range len load get list print_error exit values open get exit print_error get validate list print_error __contains__ exit keys range len exit print_error validate_customized_file get validate_customized_file validate_customized_file get validate_search_space_content exit print_error exit print_error get format print_error exit get_yml_content get format print_error exit validate_pai_config_path print_warning get parse_tuner_content parse_path print_error parse_advisor_content exit parse_assessor_content validate_pai_trial_conifg parse_time validate_annotation_content validate_common_content print version add_argument add_parser func ArgumentParser set_defaults add_subparsers convert_time_stamp_to_date rest_get text experiment_url loads check_rest_server_quick Config Experiments list isinstance get_experiment_time get_config get_experiment_status get_all_experiments keys update_experiment Experiments list print_normal get isinstance print_error print exit id get_all_experiments append remove_experiment keys update_experiment Experiments list print_normal all join isinstance print_error endswith print exit id startswith get_all_experiments append remove_experiment keys Experiments print_error exit check_experiment_id get_all_experiments Experiments print_error exit check_experiment_id get_all_experiments get str strftime Config print_normal get_config check_rest_server_quick get_config_filename Config Experiments update_experiment print_normal time str parse_ids get_config exit system strftime kill_command localtime get_all_experiments print_warning set_config Config convert_time_stamp_to_date rest_get print_error print get_config text dumps check_rest_server_quick loads get_config_filename trial_jobs_url enumerate Config trial_job_id_url print_error print get_config text trial_id check_rest_server_quick get_config_filename rest_delete Config print_error expand_annotations exit trial_id get_config_filename check_experiment_id print_warning Config convert_time_stamp_to_date rest_get print_error print get_config text dumps check_rest_server_quick experiment_url loads get_config_filename Config print_normal print get_config text dumps check_rest_server_quick loads get_config_filename print join check_output_command get_config_filename log_internal log_internal Config get print_normal format rest_get print_error get_config text exit trial_id check_rest_server_quick loads get_config_filename append trial_jobs_url Config print get_config_filename get_all_config Config join print_normal format get_config get_config_filename get_nni_installation_path join print_normal run rmtree print_normal format get join str print_normal format create_ssh_sftp_client remove_remote_directory join print_normal format HdfsClient group delete match print_warning compile Config print_error id hdfs_clean remove_experiment str list print_normal all remote_clean local_clean exit get_all_experiments input append home get format get_config eval print_warning keys Experiments join print get str format append get config update_experiment print_normal format input print_error print remote_clean exit get_yml_content eval hdfs_clean abspath get_platform_dir print_warning update_experiment Experiments list print_normal all print exit get_all_experiments append print_warning keys seconds strptime mktime update_experiment Experiments list convert_time_stamp_to_date rest_get print text exit check_rest_server_quick loads get_all_experiments append print_warning trial_jobs_url keys enumerate update_experiment print_normal format show_experiment_info system exit get_experiment_status sleep kill_command time exit set_monitor print_error Config rest_get print_error get_config text exit check_rest_server_quick type loads get_config_filename append export_data_url join print_normal format trial_dir getcwd wait file print_warning expanduser exists join install_requirements_command print_error name process_install print join list keys put post get delete check_status_url range rest_get sleep check_status_url rest_get join listdir makedirs Transport connect from_private_key_file from_transport check_environment join remove listdir rmdir print_error search group trial_id exit append get join print_normal copy_remote_directory_to_local create_ssh_sftp_client append enumerate get join print_normal print_error parse_log_path copy_data_from_remote exit append enumerate str join print_normal pid print_error get_config exit port detect_port append get_local_urls set_config Config Experiments print_normal print_error get_config call check_experiment_id get_all_experiments set_config Config Experiments join get_path_list rest_get print_error get_config text exit gettempdir check_rest_server_quick loads check_experiment_id get_all_experiments trial_jobs_url start_tensorboard_process makedirs print_warning get_config exit print_error get_json_content dumps Config rest_get print_error get_config text rest_put dumps get_query_type check_rest_server_quick experiment_url loads get_config_filename print_normal load_search_space print_error get_experiment_port validate_file filename update_experiment_profile int value print_normal print_error get_experiment_port validate_digit update_experiment_profile int value print_normal print_error get_experiment_port parse_time update_experiment_profile int value print_normal print_error validate_digit update_experiment_profile load_search_space print_error get_experiment_port validate_file validate_dispatcher filename import_data_to_restful_server Config print_error rest_post get_config import_data_url check_rest_server_quick get_config_filename items list format address append pop int remove list check_output map getpid splitlines append split check_output exit parse_nvidia_smi_result split getElementsByTagName umask parseString umask join pathSuffix list_status copyHdfsFileToLocal makedirs remove format nni_log length copy_to_local get_file_status Info join mkdirs listdir isdir copy_from_local delete exists isdir print format now value hdfs_host webhdfs_path HdfsClient pai_hdfs_host pid nnimanager_port Stdout copyHdfsDirectoryToLocal Popen copyDirectoryToHdfs getcwd exit nnimanager_ip get_hdfs_client sleep Info nni_hdfs_exp_dir format hdfs_output_dir trial_command pai_hdfs_output_dir poll nni_log RemoteLogger log_collection makedirs split get_pipelog_reader print Error format nni_log rest_post gen_send_version_url group nnimanager_port nnimanager_ip dumps Warning version Info _exit str join format nni_log basename get_hdfs_client Debug copyHdfsFileToLocal listdir start FetchThread

   **<font size=4>开发单位：鹏城实验室(PCL)，清华大学(THU)</font>** **<font size=4>特别感谢国防科技大学窦勇老师及其团队的宝贵意见和支持</font>** # <span id="head1">AIPerf Benchmark v1.0</span> ## <span id="head2"> Benchmark结构设计</span> **关于AIPerf设计理念，技术细节，以及测试结果，请参考论文：https://arxiv.org/abs/2008.07141** AIPerf Benchmark基于微软NNI开源框架，以自动化机器学习（AutoML）为负载，使用network morphism进行网络结构搜索和TPE进行超参搜索。 ## <span id="head3"> Benchmark安装说明</span> **本文用于在容器环境下运行Benchmark**

AI-secure/VeriGauge

['autonomous driving']

['SoK: Certified Robustness for Deep Neural Networks']

convex_adversarial/examples/cifar.py recurjac/task_lipschitz.py convex_adversarial/examples/fashion_mnist.py cnn_cert/setup_mnist.py cnn_cert/train_resnet.py cnn_cert/CLEVER/setup_mnist.py recurjac/bound_fastlin_fastlip.py basic/core.py cnn_cert/CLEVER/nlayer_model.py cnn_cert/CLEVER/randsphere.py models/cnn_cert_model.py basic/stats.py recurjac/activation_functions.py recurjac/activation_functions_quad.py crown_ibp/datasets.py convex_adversarial/convex_adversarial/dual_inputs.py experiments/params.py basic/milp.py crown_ibp/converter/setup_mnist.py crown_ibp/eps_scheduler.py recurjac/bound_crown_quad.py adaptor/crown_adaptor.py experiments/evaluate.py cnn_cert/CLEVER/collect_gradients.py cnn_cert/cnn_to_mlp.py crown_ibp/train.py crown_ibp/convex_adversarial/dual_network.py recurjac/main.py eran/tf_verify/eran.py recurjac/bound_spectral.py cnn_cert/ReluplexCav2017/setup_mnist.py experiments/data_analyzer.py models/test_model.py cnn_cert/CLEVER/setup_cifar.py crown_ibp/convex_adversarial/dual.py eran/testing/check_models.py crown_ibp/converter/utils.py adaptor/cnncert_adaptor.py crown_ibp/config.py convex_adversarial/convex_adversarial/dual.py models/exp_model.py cnn_cert/utils.py models/crown_ibp_model.py cnn_cert/fastlin/setup_tinyimagenet.py convex_adversarial/examples/svhn.py convex_adversarial/examples/mnist_epsilon.py recurjac/task_landscape.py recurjac/utils.py eran/tf_verify/read_net_file.py cnn_cert/fastlin/setup_cifar.py convex_adversarial/examples/trainer.py cnn_cert/pymain.py convex_adversarial/examples/primal.py eran/tf_verify/tensorflow_translator.py eran/tf_verify/deeppoly_nodes.py adaptor/basic_adaptor.py convex_adversarial/convex_adversarial/__init__.py crown_ibp/convex_adversarial/utils.py cnn_cert/fastlin/get_bounds_ours.py convex_adversarial/examples/har.py crown_ibp/pgd.py recurjac/cachefy.py eran/tf_verify/optimizer.py main.py basic/utils.py model.py basic/models.py adaptor/lpdual_adaptor.py crown_ibp/converter/setup_cifar.py recurjac/converter/setup_mnist.py cnn_cert/Attacks/l1_attack.py recurjac/converter/mnist_cifar_models.py eran/tf_verify/onnx_translator.py recurjac/converter/keras2torch.py cnn_cert/CLEVER/clever.py cnn_cert/CLEVER/estimate_gradient_norm.py recurjac/train_nlayer.py convex_adversarial/convex_adversarial/dual_network.py crown_ibp/convex_adversarial/__init__.py convex_adversarial/examples/attacks.py cnn_cert/setup_tinyimagenet.py eran/tf_verify/ai_milp.py recurjac/mnist_cifar_models.py cnn_cert/Attacks/l2_attack.py experiments/trainer.py recurjac/setup_mnist.py convex_adversarial/examples/problems.py eran/tf_verify/read_zonotope_file.py crown_ibp/model_defs_gowal.py convex_adversarial/examples/mnist.py recurjac/task_robustness.py cnn_cert/fastlin/save_nlayer_weights.py cnn_cert/CLEVER/shmemarray.py recurjac/converter/utils.py crown_ibp/converter/keras2torch.py cnn_cert/train_cnn.py cnn_cert/CLEVER/CNNModel.py cnn_cert/fastlin/get_bounds_ours_sparse.py cnn_cert/activations.py models/zoo.py recurjac/parse_landscape.py constants.py cnn_cert/Attacks/li_attack.py crown_ibp/argparser.py cnn_cert/train_lenet.py models/recurjac_model.py crown_ibp/eval.py recurjac/setup_cifar.py crown_ibp/converter/mnist_cifar_models.py cnn_cert/cnn_bounds_full_core.py convex_adversarial/convex_adversarial/utils.py eran/tf_verify/analyzer.py convex_adversarial/examples/cifar_evaluate.py basic/intervalbound.py cnn_cert/setup_cifar.py basic/model_prepare.py recurjac/bound_crown.py basic/percysdp.py cnn_cert/cnn_bounds_full.py crown_ibp/converter.py crown_ibp/convex_adversarial/dual_inputs.py convex_adversarial/convex_adversarial/dual_layers.py recurjac/bound_interval.py adaptor/recurjac_adaptor.py cnn_cert/Attacks/cw_attack.py recurjac/bound_recurjac.py cnn_cert/fastlin/main_sparse.py cnn_cert/fastlin/utils.py cnn_cert/train_nlayer.py eran/tf_verify/config.py cnn_cert/ReluplexCav2017/to_nnet.py basic/lowerbound.py cnn_cert/fastlin/get_bounds_others.py eran/tf_verify/__main__.py recurjac/parse_lipschitz.py recurjac/converter/torch2keras.py eran/tf_verify/constraints.py adaptor/eran_adaptor.py recurjac/bound_base.py crown_ibp/converter/torch2keras.py recurjac/converter/setup_cifar.py adaptor/adaptor.py basic/components.py cnn_cert/fastlin/main.py convex_adversarial/examples/runtime.py cnn_cert/CLEVER/process_log.py crown_ibp/attack.py crown_ibp/bound_layers.py eran/data/create_zonotope.py experiments/model_pretransform.py eran/tf_verify/eranlayers.py eran/tf_verify/deepzono_nodes.py cnn_cert/CLEVER/defense.py cnn_cert/fastlin/setup_mnist.py eran/tf_verify/krelu.py experiments/model_summary.py datasets.py crown_ibp/convex_adversarial/dual_layers.py crown_ibp/model_defs.py basic/fastmilp.py _mnist _imagenet get_num_classes get_dataset _cifar10 NormalizeLayer get_input_shape pr load_model Adaptor BasicAdaptor RealAdaptorBase CWAdaptor FastLinIBPAdaptor PGDAdaptor VerifierAdaptor FastMILPAdaptor PercySDPAdaptor MILPAdaptor IBPAdaptor CleanAdaptor FazlybSDPAdaptor CNNCertBase fn FastLinSparseAdaptor LPAllAdaptor CNNCertAdaptor check_consistency sequential_torch2keras FullCrownAdaptor CrownAdaptorBase CrownIBPAdaptor IBPAdaptor AI2Adaptor RefineZonoAdaptor KReluAdaptor DeepZonoAdaptor ERANBase DeepPolyAdaptor init_domain RefinePolyAdaptor ZicoDualAdaptor FastLipAdaptor RecurJacAdaptor FastLinAdaptor SpectralAdaptor torch2keras RecurJacBase RecurBaseModel FastMILPVerifier FastIntervalBound IntervalFastLinBound BoundCalculator IntervalBound FastLinBound MILPVerifier to_var load_model LinfPGDAttack PercySDP data_loaders atand ada_linear_bounds atan_linear_bounds sigmoidut tanhlt tanhd tanhut sigmoidd tanh_linear_bounds tanhid sigmoidid sigmoidlt relu_linear_bounds atanut tanh atanid sigmoid atanlt atan sigmoid_linear_bounds UL_basic_block_2_bound UL_conv_bound UL_basic_block_bound UL_pool_bound Model UL_relu_bound compute_bounds find_output_bounds loss warmup run pool conv_full lower_bound_conv pool_linear_bounds upper_bound_conv fn conv_bound lower_bound_pool conv compute_bounds run upper_bound_pool find_output_bounds CNNModel warmup conv_bound_full fn get_weights printlog convert printlog run_all_general run_LP run_cnn command run_CLEVER run_global run run_attack run_all_relu TwoLayerCIFARModel CIFARModel CIFAR load_batch MNIST extract_data TwoLayerMNISTModel extract_labels MNISTModel MadryMNISTModel load_images tinyImagenet train train train_cnn_7layer Residual ResidualStart2 train ResidualStart Residual2 show l0_dist generate_data l2_dist l1_dist linf_dist loss cw_attack EADL1 CarliniL2 CarliniLi get_lipschitz_estimate fmin_with_reg parse_filename get_best_weibull_fit fit_and_test plot_weibull CNNModel collect_gradients defend_png defend_jpeg defend_tv bregman defend_crop make_defend_quilt defend_reduce defend_none fn EstimateLipschitz NLayerModel readDebugLog2array gen_table l1_samples randsphere randsign l2_samples linf_samples TwoLayerCIFARModel CIFARModel CIFAR load_batch MNIST extract_data TwoLayerMNISTModel extract_labels MNISTModel address_of_buffer NpShmemArray ShmemRawArray ShmemBufferWrapper get_layer_bound_LP spectral_bound fast_compute_max_grad_norm_zfw compute_max_grad_norm fast_compute_max_grad_norm_2layer_zfw fast_compute_max_grad_norm_2layer_z fast_compute_max_grad_norm_z compute_worst_bound ReLU fast_compute_max_grad_norm_2layer fast_compute_max_grad_norm_2layer_next fast_compute_max_grad_norm_2layer_fw inc_counter get_weights_list get_layer_bound fast_compute_max_grad_norm fast_compute_max_grad_norm_2layer_next_fw fast_compute_max_grad_norm_2layer_next_zfw init_layer_bound_relax_matrix_huan fast_compute_max_grad_norm_fw fast_compute_max_grad_norm_2layer_next_linear fast_compute_max_grad_norm_linear fast_compute_max_grad_norm_2layer_next_z get_layer_bound_relax_matrix_huan_optimized compute_worst_bound_multi get_layer_bound_relax_matrix_huan_optimized compute_worst_bound init_layer_bound_relax_matrix_huan ReLU get_weights_list get_layer_bound NLayerModel loss CNNModel TwoLayerCIFARModel CIFARModel CIFAR load_batch MNIST extract_data TwoLayerMNISTModel extract_labels MNISTModel MadryMNISTModel load_images tinyImagenet show l0_dist generate_data l2_dist l1_dist linf_dist MNIST extract_data TwoLayerMNISTModel extract_labels MNISTModel MadryMNISTModel fn get_weights_biases nnet DualObject DualLayer InfBallProjBounded InfBall select_input InfBallBounded L2Ball InfBallProj L2BallProj DualLinear conv_transpose2d Identity DualDense DualBatchNorm2d select_layer conv2d DualReshape DualConv2d unbatch DualReLU batch DualReLUProj robust_loss_parallel RobustBounds DualNetBounds robust_loss InputSequential DualNetwork get_epsilon epsilon_from_model DenseSequential GR Dense GL p_lower full_bias p_upper _fgs pgd mean _pgd fgs attack Flatten select_model select_model select_model f har_500_250_model cifar_model_resnet cifar_model mnist_loaders mnist_model_large Flatten mnist_model_wide mnist_500 har_500_model har_resnet_model replace_10_with_0 svhn_model cifar_loaders cifar_model_large har_500_250_100_model fashion_mnist_loaders args2kwargs argparser_evaluate svhn_loaders model_wide mnist_model_deep argparser mnist_model model_deep har_loaders Meter gpu_mem train_baseline evaluate_baseline sampler_robust_cascade AverageMeter evaluate_robust_cascade evaluate_robust evaluate_madry train_robust robust_loss_cascade train_madry isfloat isint argparser BoundSequential BoundDataParallel BoundFlatten BoundReLU BoundConv2d BoundLinear config_dataloader config_modelloader_and_convert2mlp get_file_close get_path config_modelloader update_dict load_config get_model_config main get_stats mnist_loaders svhn_loaders cifar_loaders EpsilonScheduler main model_cnn_4layer convert_conv2d_dense DenseConv2d model_cnn_1layer model_mlp_any model_mlp_uniform load_checkpoint_to_mlpany save_checkpoint model_cnn_10layer model_cnn_3layer_fixed model_cnn_2layer IBP_large IBP_debug pgd output mean _pgd attack main AverageMeter Logger Train Flatten NLayerModel get_model_meta_real get_model_meta CIFAR load_batch MNIST extract_data extract_labels model_mlp_any model_mlp_uniform Flatten show l0_dist generate_data l2_dist l1_dist linf_dist binary_search DualObject DualLayer InfBallProjBounded InfBall select_input InfBallBounded L2Ball InfBallProj L2BallProj DualLinear conv_transpose2d Identity DualFeatureMask2D DualDense DualBatchNorm2d select_layer conv2d DualReshape DualConv2d unbatch DualReLU batch DualReLUProj robust_loss_parallel RobustBounds DualNetBounds robust_loss InputSequential DualNetwork get_epsilon epsilon_from_model DenseSequential GR Dense GL p_lower full_bias p_upper get_tests normalize get_out_tensors handle_affine get_bounds_for_layer_with_milp handle_conv create_model verify_network_with_milp solver_call handle_residual Cache handle_relu handle_maxpool Analyzer layers config Device get_constraints_for_dominant_label label_index clean_string get_constraints_from_file DeeppolyMaxpool DeeppolyGather calc_bounds DeeppolyNode DeeppolyReluNodeFirst DeeppolyTanhNodeIntermediate DeeppolyInput DeeppolySigmoidNodeIntermediate DeeppolySubNodeIntermediate DeeppolyMulNodeIntermediate DeeppolyTanhNodeLast DeeppolyConv2dNodeFirst DeeppolyMulNodeFirst DeeppolyConv2dNodeIntermediate DeeppolyReluNodeIntermediate add_input_output_information_deeppoly DeeppolySubNodeFirst DeeppolyReluNodeLast DeeppolySigmoidNodeFirst DeeppolyTanhNodeFirst DeeppolyResadd DeeppolySigmoidNodeLast DeepzonoAffine DeepzonoMatmul DeepzonoNonlinearity DeepzonoTanh add_bounds DeepzonoSigmoid DeepzonoSub DeepzonoInputZonotope add_dimensions DeepzonoConvbias remove_dimensions DeepzonoResadd DeepzonoRelu add_input_output_information refine_relu_with_solver_bounds DeepzonoAdd DeepzonoMaxpool DeepzonoDuplicate DeepzonoConv DeepzonoMul DeepzonoGather get_xpp DeepzonoInput ERAN eran_input eran_dense eran_conv2d_without_activation eran_maxpool eran_activation eran_resnet_dense tensorshape_to_intlist eran_affine eran_resnet_conv2d eran_reshape eran_conv2d generate_linexpr0 make_krelu_obj grouping_heuristic encode_krelu_cons Krelu prepare_model ONNXTranslator reshape_nhwc nchw_to_nhwc onnxshape_to_intlist Optimizer permutation product numel runRepl parseVec read_tensorflow_net extract_mean myConst read_onnx_net extract_std read_zonotope calculate_padding tensorshape_to_intlist TFTranslator isnetworkfile get_tests denormalize parse_input_box normalize_poly normalize init_domain str2bool show_ascii_spec texify_percentage nice_print read_verify_data read_radius_data verify_texify texify_entire radius_texify texify_radius read_file texify_time _async_raise terminate verify_wrapper radius_wrapper try_load_weight nicenum struc_summary tex_output eliminate_redundent param_summary main calc_tot_torun load_from_tf load_cnn_cert_model sequential_keras2torch crown_ibp_mnist_cnn_3layer_fixed_kernel_5_width_16 load_model ibp_cifar_large ibp_mnist_large crown_ibp_mnist_cnn_2layer_width_2 crown_ibp_cifar_cnn_3layer_fixed_kernel_3_width_16 crown_ibp_cifar_cnn_2layer_width_2 mnist_conv_small in_cells two_layer_fc20 conv_super three_layer_fc100 mnist_conv_large mnist_conv_medium seven_layer_fc1024 cifar_conv_medium cifar_conv_large cifar_conv_small try_load_weight get_input_shape load_keras_model abstract_load_keras_model get_normalize_layer test_mnist_tiny test_mnist test_cifar10 test_cifar10_tiny model_cnn_4layer IBP_large model_cnn_1layer model_cnn_2layer model_mlp_any IBP_debug model_mlp_uniform cifar_model mnist_model model_cnn_10layer model_cnn_3layer_fixed mnist_model_tiny cifar_model_tiny Flatten relu_ub_p leaky_relu_lb_p act_sigmoid_d act_sigmoid relu_ub_n leaky_relu_ub_n general_ub_pn_scalar find_d_LB act_arctan plot_line leaky_relu_lb_n general_lb_n leaky_relu_ub_p act_tanh general_ub_n relu_ub_pn leaky_relu_ub_pn find_d_UB act_arctan_d general_ub_p relu_lb_n general_lb_pn general_lb_pn_scalar general_lb_p act_tanh_d relu_lb_pn leaky_relu_lb_pn relu_lb_p general_ub_pn get_lower_quad_parameterized_lgtu get_lower_quad_parameterized get_area plot_parameterized get_best_upper_quad get_upper_quad_parameterized get_best_lower_quad get_lower_quad_parameterized_lltu compute_bounds_integral compute_bounds ReLU get_weights_list myprint crown_adaptive_bound init_crown_bounds get_leaky_relu_bounds compile_crown_bounds get_general_bounds crown_general_bound get_relu_bounds proj_l2 proj_l1 proj_li crown_quad_bound get_layer_bound_quad_both fastlip_2layer_leaky fastlip_2layer fastlin_bound fastlip_bound fastlip_nplus1_layer init_fastlin_bounds fastlip_leaky_bound fastlip_nplus1_layer_leaky interval_bound recurjac_bound_2layer sigmoid_grad_bounds recurjac_bound_next_with_grad_bounds compile_recurjac_bounds recurjac_bound_backward_improved_1x1 recurjac_bound_wrapper recurjac_bound_forward_improved get_final_bound recurjac_bound_launcher recurjac_bound relu_grad_bounds last_layer_abs_sum construct_function form_grad_bounds recurjac_bound_next recurjac_bound_backward_improved recurjac_bound_forward_improved_1x1 spectral_bound add_cache remove_cache NLayerModel get_model_meta_real get_model_meta parse_result_line get_fig_info parse_single gen_title gen_legend gen_filename parse_single get_color parse_result_line get_linestyle gen_title get_fig_info CIFAR load_batch MNIST extract_data extract_labels task task task train show l0_dist generate_data l2_dist l1_dist linf_dist binary_search Flatten NLayerModel get_model_meta_real get_model_meta CIFAR load_batch MNIST extract_data extract_labels model_mlp_any model_mlp_uniform Flatten show l0_dist generate_data l2_dist l1_dist linf_dist binary_search ToTensor Compose list Compose ToTensor extend append join list ToTensor Compose extend Normalize append Graph Session norm print transpose random eval expand_dims numpy predict list set is_available cuda MNIST DataLoader Normalize CIFAR10 actd range act actd range act actd range act actd range act actd range act actd range act zeros range shape zeros range shape act actut shape actlt zeros actd range act actut shape actlt zeros actd range act actut shape actlt zeros actd range minimum maximum copy zeros range linear_bounds asarray UL_conv_bound shape UL_relu_bound linear_bounds asarray UL_conv_bound shape UL_relu_bound minimum zeros_like maximum copy range minimum maximum copy zeros range UL_basic_block_2_bound asarray linear_bounds ones reshape UL_pool_bound UL_basic_block_bound pool_linear_bounds copy UL_conv_bound shape UL_relu_bound zeros range conv_bound_full append compute_bounds range len list asarray print reshape sizes ascontiguousarray pads strides types func append clear_session pads set_random_seed flatten argmax log seed str list exp load_model squeeze Model types append range asarray inf CIFAR format generate_data sizes astype MNIST minimum time print reshape float32 strides maximum tinyImagenet find_output_bounds warmup len zeros range zeros max range abs sqrt shape conv zeros sum max range zeros range conv_full abs sqrt shape zeros sum max range minimum linear_bounds conv_full maximum shape zeros range minimum linear_bounds conv_full maximum shape zeros range inf min astype float32 where dot zeros sum max range len minimum conv_full pool_linear_bounds maximum shape zeros range minimum conv_full pool_linear_bounds maximum shape zeros range lower_bound_conv maximum upper_bound_conv lower_bound_pool upper_bound_pool tuple conv_bound fn CNNModel recompile print load_model print append zeros prod range MNIST printlog format CIFAR print compile Lambda Sequential len SGD add shape Dense save get_weights range Flatten set_weights command printlog split printlog format run_cnn run append len append run_cnn printlog format printlog format run append len collect_gradients printlog append format printlog format run append len append printlog cw_attack format read transpose fromstring append range seed join dtype permutation format print transpose float32 zeros listdir array open Sequential train_labels save train_data Flatten len Adam add format Lambda Dense load_weights zip compile print summary Conv2D BatchNormalization fit relu SGD shape MaxPooling2D Sequential SGD MaxPooling2D train_labels save train_data Flatten Adam add shape Dropout format Dense load_weights compile print fit summary Conv2D len Model Input fromarray join print squeeze flatten around save range flatten argmax str list predictor squeeze argmin len savetxt append range format set sample join print extend eye randint array makedirs seed MNIST time CIFAR load_model generate_data norm_fn set_random_seed tinyImagenet average attack kstest fit format subplots plot xlabel min close ylabel pdf title hist savefig linspace legend max nanmax format partial print map copy index fmin zip append amin amax print get_best_weibull_fit int basename splitext split get_lipschitz_estimate set_random_seed Pool seed sorted squeeze append format glob close ConfigProto flush join print parse_filename system EstimateLipschitz loadmat len uint8 astype float32 fromarray uint8 BytesIO astype float32 save fromarray uint8 BytesIO astype float32 save fromarray asarray inf shape sqrt zeros range uniform load reshape float32 placeholder matmul shape top_k to_float strip close open append split format write randint sum randn sort uniform empty reshape reshape l2_samples NpShmemArray float32 shape linf_samples transform_new range l1_samples clip get isinstance from_address get_address ShmemBufferWrapper __init__ sizeof len int itemsize ShmemRawArray c_char prod int norm format inf print ones shape zip append expand_dims range len addVar setParam MAXIMIZE str LinExpr Model objVal append range update format astype empty_like MINIMIZE time optimize print addConstr float32 setObjective reset len format print transpose U ascontiguousarray shape append get_weights enumerate sum norm empty_like dot abs max range ones empty range len sum norm zeros_like astype maximum float32 copy empty_like dot abs max range zeros_like astype float32 maximum dot zeros range zeros fast_compute_max_grad_norm_2layer range zeros_like maximum zeros abs fast_compute_max_grad_norm_2layer range fast_compute_max_grad_norm_2layer_next zeros_like astype float32 maximum dot zeros range astype float32 maximum dot zeros range list zeros_like fast_compute_max_grad_norm_2layer_fw maximum fast_compute_max_grad_norm_2layer_next_fw zeros abs range minimum zeros maximum zeros_like minimum zeros maximum range ones maximum fast_compute_max_grad_norm_2layer_next_z fast_compute_max_grad_norm_2layer_z zeros abs range minimum zeros maximum zeros_like minimum zeros maximum range list ones maximum fast_compute_max_grad_norm_2layer_zfw fast_compute_max_grad_norm_2layer_next_zfw zeros abs range zeros range fast_compute_max_grad_norm_2layer_next_linear zeros_like abs maximum dot sqrt zeros sum range len norm inc_counter zeros empty range format print compute_worst_bound any linspace range tuple compute_max_grad_norm ReLU ones append expand_dims range get_layer_bound format inf astype fast_compute_max_grad_norm empty_like init_layer_bound_relax_matrix_huan ord get_layer_bound_LP print min float32 get_layer_bound_relax_matrix_huan_optimized zeros len csr_matrix A shape zeros empty A csr_matrix multiply csr_matrix T load_model append range len append get_weights_biases len list detach isinstance zip print Conv2d ReLU BatchNorm2d Linear append append size item DataParallel set_start unsqueeze ReLU cuda is_cuda l append bounds size select_layer reversed InputSequential zip item enumerate T isinstance Variable any out_features int isinstance numel Conv2d unsqueeze Linear get_epsilon isinstance print numel unsqueeze append sum l time max p_upper p_lower data model backward Variable size zero_grad Adam sign sum data model Variable backward size clamp zero_grad Adam sign sum range format atk print squeeze mean parameters append robust_loss_batch enumerate cuda format print model_factor mnist_loaders unsqueeze DualNetBounds Sequential Conv2d ReLU Flatten Linear Sequential group ReLU Flatten Linear MNIST DataLoader MNIST DataLoader Sequential ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear DataLoader SVHN cuda TensorDataset float long DataLoader Sequential ReLU Linear Sequential ReLU Linear Sequential ReLU Linear Sequential DenseSequential Dense ReLU Linear DataLoader Normalize CIFAR10 isinstance Sequential out_channels Conv2d normal_ sqrt modules zero_ ReLU Flatten Linear isinstance Sequential out_channels Conv2d normal_ sqrt modules zero_ ReLU Flatten Linear isinstance block DenseSequential extend out_channels Conv2d normal_ sqrt modules zero_ range sorted format print add_argument prefix ArgumentParser vars parse_args epsilon cuda_ids format print add_argument ArgumentParser parse_args cuda_ids new_query clip_grad_norm_ zero_grad robust_loss squeeze update format size item flush enumerate time backward Variable print AverageMeter parameters empty_cache train step len update time format model Variable print squeeze AverageMeter set_grad_enabled size robust_loss eval item empty_cache sum enumerate flush len update time format model Variable backward size AverageMeter zero_grad step print train sum enumerate flush len update time format model Variable print size AverageMeter eval sum enumerate flush len data model zero_grad sign Adam sum range update format size item flush enumerate time backward Variable clamp print AverageMeter train step len update time format model Variable print size AverageMeter _pgd eval item sum enumerate flush len data model size type_as rl item append float sum enumerate format print Variable set_grad_enabled SubsetRandomSampler DataLoader cat robust_loss_cascade dataset cuda enumerate append len update time format Variable print squeeze AverageMeter set_grad_enabled size eval item robust_loss_cascade empty_cache enumerate flush len float int seed int isint isfloat overrides seterr manual_seed float split glob sorted format print get items list format isinstance print config str format deepcopy replace isinstance list print items overrides_dict model_subset update_dict append range path_prefix get_file_close join makedirs load list model_class isinstance print get_path keys import_module getattr load_state_dict append cuda load convert_conv2d_dense list model_class format isinstance print get_path load_checkpoint_to_mlpany dense_m import_module getattr load_state_dict save_checkpoint append zeros keys load_config config_modelloader_and_convert2mlp view size sqrt dataset len list Subset dataset range list Subset range get_stats list print Subset Normalize range config_dataloader Logger argmax max log cuda open config_modelloader argmin append update format get_path mean zip float deepcopy print convert min median array append Linear ReLU Sequential Conv2d ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear children list DenseConv2d isinstance kernel_size out_channels in_channels Conv2d bias __class__ append weight children list isinstance dense_w bias dense_bias OrderedDict save weight Linear join format replace endswith print model_mlp_any shape load_state_dict append state_dict Sequential Conv2d ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear enumerate view output get_eps batch_size model zero_grad scatter_ unsqueeze tensor abs max log is_cuda cuda view new f step expand scatter sum range detach get update format inf LongTensor size mean eval item DualNetwork enumerate int time backward reshape AverageMeter min named_parameters cpu zeros train numpy std get_eps BoundDataParallel batch_size SGD MultiStepLR save Train dataset str StepLR Adam ceil EpsilonScheduler get_lr range get inf int join time update_dict parameters step std len print get_model_meta_real model_loader add set loads append to_json enumerate print format cond str read format reader open range len addTerms str addConstant addVar LinExpr addConstr EQUAL append prod range len str GREATER_EQUAL LESS_EQUAL addVar LinExpr addConstr addGenConstrIndicator EQUAL append float range len addTerms str addConstant addVar LinExpr addConstr EQUAL append range len str addConstant addVar LinExpr addConstr EQUAL append range len addTerms str cons addConstant GREATER_EQUAL LESS_EQUAL addVar LinExpr addConstr addGenConstrIndicator varsid EQUAL append max range enumerate len addConstant maxpool_counter addVar setParam abs addTerms str handle_affine LinExpr FeasibilityTol Model append prod range handle_conv EQUAL conv_counter residual_counter ffn_counter addConstr handle_residual handle_relu handle_maxpool len max optimize LinExpr min copy setObjective reset MINIMIZE MAXIMIZE flush update sorted calc_layerno create_model zip print numlayer Threads reset TimeLimit numproc setParam append zeros range enumerate len optimize create_model LinExpr numlayer setObjective setParam MINIMIZE len CPU append range int clean_string readlines len label_index startswith append range split calc_layerno elina_interval_array_free get_num_neurons_in_layer encode_krelu_cons box_for_layer append reduce elina_dimchange_alloc elina_dimchange_free elina_abstract0_add_dimensions elina_dimchange_init range elina_dimchange_alloc elina_dimchange_free elina_dimchange_init range elina_abstract0_remove_dimensions reduce calc_layerno abstract_information get_bounds_for_layer_with_milp specUB numlayer relu_zono specLB encode_krelu_cons relu_zono_layerwise append relu_zono_refined range elina_interval_array_free realdim elina_abstract0_to_box append intdim elina_abstract0_dimension Variable reshape reduce matmul tensorshape_to_intlist shape Variable int conv2d sigmoid tanh relu append append eran_affine eran_conv2d_without_activation int eran_conv2d_without_activation add eran_conv2d range eran_dense reduce add tensorshape_to_intlist shape eran_affine eran_reshape range ElinaDim coeff cst pointer elina_scalar_set_double ELINA_LINEXPR_SPARSE elina_linexpr0_alloc zip enumerate scalar len sorted pop ElinaDim sorted asarray int combinations time str print debug remove_dimensions min grouping_heuristic sqrt append add_dimensions len list reshape transpose prod initializer floor to_array attribute list name multiply node add shape i input append expand_dims ceil prod range onnxshape_to_intlist concatenate ints insert subtract copy take float s int t zeros nchw_to_nhwc len replace float64 search group split zeros range len float64 search group split zeros range len zeros range float64 concat myConst bias_add open transpose placeholder matmul add runRepl conv2d readline relu extract_mean extract_std tanh print reshape max_pool sigmoid parseVec load node check_model int read array split int max splitext product replace findall append split print range zeros range len zeros range len int list float dict append bool format print copy range enumerate print str format enumerate print str format enumerate list dict read_file append sum keys exists enumerate list dict read_file append sum keys exists enumerate print format enumerate verify time int verify time calc_radius PyThreadState_SetAsyncExc type c_long py_object _async_raise ident load print load_state_dict str remove sum isinstance model print apply OrderedDict abs prod len isinstance print refresh ReLU Linear range len enumerate print join data CallableModelWrapper set_grad_enabled zero_grad sign getpass DataLoader ReLU tf_model try_load_weight get_input_shape Session list run convert_pytorch_model_to_tf placeholder get_dataset load_state_dict generate sum save_with_configs size close eval item flush Linear load enumerate isinstance backward Variable clamp makedirs float32 ProjectedGradientDescent empty_cache m Tensor train load_model layers rate kernel_size MaxPool2D Sigmoid ReLU tensor Flatten Parameter list transpose MaxPool2d append get_weights BatchNormalization Lambda Tanh Dense __name__ Linear T input_shape isinstance pool_size strides output_shape Conv2d Conv2D BatchNorm2d Tensor Dropout join print load_from_tf cuda summary sequential_keras2torch load list isinstance print load_state_dict keys cuda model_cnn_2layer load_model model_cnn_3layer_fixed cuda load_model get_normalize_layer load_model Sequential cuda model_cnn_2layer get_normalize_layer load_model Sequential model_cnn_3layer_fixed cuda IBP_large cuda load_model IBP_large get_normalize_layer load_model Sequential cuda get_input_shape get_normalize_layer Sequential in_cells Sequential ReLU Flatten Linear in_cells Sequential ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear isinstance Sequential out_channels Conv2d normal_ sqrt modules zero_ ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear isinstance Sequential out_channels Conv2d normal_ sqrt modules zero_ ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear isinstance Sequential out_channels Conv2d normal_ sqrt modules zero_ ReLU Flatten Linear Sequential Linear Conv2d ReLU Flatten get_input_shape in_cells Sequential ReLU Flatten Linear layers rate Sequential copy_ ReLU Activation cuda Flatten list load_model append get_weights LeakyReLU range copy Tanh Dense alpha Linear isinstance print Tensor Dropout get_input_shape load get_normalize_layer Sequential load_state_dict cuda load get_normalize_layer Sequential load_state_dict cuda load get_normalize_layer Sequential load_state_dict cuda load get_normalize_layer Sequential load_state_dict cuda Sequential Conv2d ReLU Flatten Linear Sequential Conv2d ReLU Flatten Linear zeros_like full zeros_like len cosh dfunc logical_not func empty_like dfunc func dfunc func dfunc logical_not func empty_like empty_like func find_d_UB range len find_d_LB empty_like func range len func find_d_UB func find_d_LB range diff range diff plot print get_area format print get_area format get_upper_quad_parameterized func linspace plot norm get format print any compute_bounds linspace range minimum print set_printoptions eval input sum init_crown_bounds compile_recurjac_bounds tuple recurjac_bound_wrapper ReLU crown_quad_bound max crown_adaptive_bound compile_crown_bounds crown_general_bound append expand_dims sum format inf astype float flush pop minimum interval_bound print fastlin_bound fastlip_bound min float32 init_fastlin_bounds fastlip_leaky_bound myprint len ub_pn lb_pn lb_p lb_n ub_n ub_p ub_pn lb_pn lb_p lb_n ub_n ub_p ub_pn lb_pn lb_p lb_n ub_n ub_p ones empty range len format print exec globals compile interval_bound zeros_like empty_like copy dot get_bounds range interval_bound zeros_like astype maximum float32 copy empty_like dot range norm arange cumsum float32 empty_like maximum abs len zeros_like proj_li grad_ub f_qp_lb range inf astype empty_like grad_lb f_qp_ub T norm print maximum float32 dot proj_l2 proj_l1 get_best_lower_quad len zeros_like get_best_lower_quad get_best_upper_quad grad_ub f_qp_lb range inf astype empty_like grad_lb f_qp_ub T norm print float32 maximum dot proj len ones empty range len interval_bound zeros_like astype maximum float32 copy empty_like dot range zeros_like astype float32 maximum dot zeros range zeros range zeros_like sum fastlip_2layer logical_and fastlip_nplus1_layer maximum zeros abs range zeros range zeros_like zeros_like print astype float32 dot zeros range fastlip_2layer_leaky print abs logical_and maximum fastlip_nplus1_layer_leaky zeros sum range dot norm range empty_like zeros prange range zeros_like zeros_like prange get_grad_bounds zeros range zeros prange range zeros_like recurjac_bound_forward_improved_1x1_LB zeros_like dot recurjac_bound_forward_improved_1x1_UB expand_dims range zeros_like recurjac_bound_backward_improved_1x1_UB reshape recurjac_bound_backward_improved_1x1_LB dot expand_dims range recurjac_bound_forward_improved_1x1_LB zeros_like prange dot recurjac_bound_forward_improved_1x1_UB zeros expand_dims range zeros_like recurjac_bound_backward_improved_1x1_UB reshape prange recurjac_bound_backward_improved_1x1_LB dot zeros expand_dims range list zeros_like abs tuple recurjac_bound_backward_improved_1x1_UB maximum logical_not recurjac_bound logical_or empty recurjac_bound_next append zeros expand_dims sum range len join getsource replace endswith exec lstrip splitlines range globals __name__ compile len exec format globals compile recurjac_bound_next_with_grad_bounds print prange logical_and recurjac_bound_launcher construct_function get_grad_bounds append sum get_grad_bounds range len zeros_like tuple empty_like recurjac_bound form_grad_bounds append zeros range len recurjac_bound_2layer list recurjac_bound_forward_improved recurjac_bound_backward_improved range len norm abs maximum last_layer_abs_sum empty max range strip replace append float strip split parse_result_line startswith enumerate int split upper ModelCheckpoint model NLayerModel

AI-secure/VeriGauge

AIEMMU/MRI_Prostate

['medical image segmentation', 'medical diagnosis']

['Deep learning in magnetic resonance prostate segmentation: A review and a new perspective']

app/contour.py app/predictor.py app/viewer.py app/hooks/hook-fastprogress.py app/hooks/hook-PySide2.QtWebEngineWidgets.py app/hooks/hook-torchvision.py app/ui.py app/displayView.py app/utils.py app/displayViewModel.py app/dicom.py Contour dicom MainWindow DisplayViewModel predictor get_y enableButtons listify loadContour setify getPixmap saveContour distance SubScene find_nearest Viewer get_relative_path_if_possible isinstance uint8 astype shape stack QPixmap Format_RGB888 QImage listify setEnabled savetxt abs sorted asarray pyqtSignal QPoint Path

# Automatic Prostate segmentation from MR images This provides some deep Learning code for training a deep learning model for segmentation from Magnetic Resonance(MR) Images. There is also a pyqt software avaialble to allow users to explore the use of the deep learning model that was trained as part of this project. More details on the model can be found at [here](https://arxiv.org/abs/2011.07795)  ## Dependencies Before running the python software, you need install its dependecies, which relies on Pytorch, FastAi, and PyDicom.You can install the dependences from the command line using the following command: ``` pip install -r requirements.txt ``` ## Usage The segmentation model file can be downloaded from [here](https://www.dropbox.com/s/a2rwhy29wx9s448/export_orig.pkl?dl=0)

AIS-Bonn/FreqNet

['video prediction']

['Frequency Domain Transformer Networks for Video Prediction']

BBall/data_dynamic.py MMNIST/data_dynamic.py show_single_V bounce_n ar bounce_vec bounce_mat sigmoid new_speeds unsigmoid show_A matricize show_V show_single_V bounce_n ar bounce_vec bounce_mat sigmoid new_speeds unsigmoid show_A matricize show_V int norm randn transpose rand dot new_speeds zeros abs array range ar meshgrid zeros array range matricize bounce_n array matricize bounce_n array show int sqrt reshape show int print reshape sqrt range len show range len

# Frequency Domain Transformer Networks for Video Prediction If you use the code for your research paper, please cite the following paper: <p> Hafez Farazi<b></b>, and Sven Behnke:<br> <a href="https://arxiv.org/pdf/1903.00271.pdf"><u>Frequency Domain Transformer Networks for Video Prediction</u></a>&nbsp;<a href="https://arxiv.org/pdf/1903.00271.pdf">[PDF]</a> <a href="http://www.ais.uni-bonn.de/~hfarazi/papers/FreqNet.bib">[BIB]</a><br> In Proceedings of 27th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN), Bruges, Belgium, April 2019.<br><b></b><br> </p>

AIS-Bonn/LocDepVideoPrediction

['video prediction']

['Location Dependency in Video Prediction']

data_dynamic.py show_single_V bounce_n ar bounce_vec bounce_mat sigmoid new_speeds unsigmoid show_A matricize show_V int norm randn transpose rand dot new_speeds zeros abs array range ar meshgrid zeros array range matricize bounce_n array matricize bounce_n array show int sqrt reshape show int print reshape sqrt range len show range len

# LocDepVideoPrediction If you use the code for your research paper, please cite the following paper: <p> Niloofar Azizi<b>*</b>, Hafez Farazi<b>*</b>, and Sven Behnke:<br> <a href="http://www.ais.uni-bonn.de/~hfarazi/papers/LocDep.pdf"><u>Location Dependency in Video Prediction</u></a>&nbsp; <a href="http://www.ais.uni-bonn.de/~hfarazi/papers/LocDep.pdf">[PDF]</a>&nbsp; <a href="http://www.ais.uni-bonn.de/~hfarazi/papers/LocDep.bib">[BIB]</a>&nbsp; <a href="https://github.com/AIS-Bonn/LocDepVideoPrediction">[CODE]</a><br> Accepted for 27th International Conference on Artificial Neural Networks (ICANN), Rhodes, Greece, to appear October 2018.<br><b> *: Both authors contributed equally to this work.</b><br> </p>

ALBERT-Inc/blog_ssap

['instance segmentation', 'semantic segmentation']

['SSAP: Single-Shot Instance Segmentation With Affinity Pyramid']

src/loss.py src/graph_partition.py src/SSAP.py src/mydatasets.py greedy_additive calc_js_div Edge make_ins_seg Partition calc_loss l2_loss focal_loss preprocess Mydatasets Up DoubleConv Down OutConv SSAP list heapify contract heappop removed heappush exp clip sum log greedy_additive sorted calc_js_div list ones nodes where log randint Edge repeat array append zeros argmax Partition range len float ones_like where l2_loss focal_loss tile

# SSAP再現実装公開用リポジトリ [SSAP [Proposal-freeなInstance Segmentation手法] の紹介と実験](https://blog.albert2005.co.jp/2020/08/18/ssap/)で使用したプログラムです. ## ファイル構成について ``` ├─ README.md # プロジェクトの説明 ├─ requirements.txt # すべてのPythonプログラムのベースとなるPythonパッケージ ├─ data/ # データフォルダ │ ├─ train2014/ # COCO2014の学習データ │ ├─ val2014/ # COCO2014の評価データ │ ├─ annotations/ # COCO2014のannotationデータ

AMLab-Amsterdam/CEVAE

['causal inference']

['Causal Effect Inference with Deep Latent-Variable Models']

cevae_ihdp.py utils.py datasets.py evaluation.py IHDP Evaluator get_y0_y1 fc_net format print write mean range flush

# CEVAE This repository contains the code for the Causal Effect Variational Autoencoder (CEVAE) model as developed at [1]. This code is provided as is and will not be updated / maintained. Sample experiment --- To perform a sample run of CEVAE on 10 replications of the Infant Health and Development Program (IHDP) dataset just type: `python cevae_ihdp.py` Other datasets

AMLab-Amsterdam/SEVDL_MGP

['gaussian processes']

['Structured and Efficient Variational Deep Learning with Matrix Gaussian Posteriors']

mnist_classification.py optimizer.py nn_utils.py layers.py yacht_regression.py matrix_layers.py VMGNet.py sample_mgaus sample_gauss sample_mult_noise add_bias kldiv_gamma sample_mgaus2 Layer MatrixGaussDiagLayerLearnP MatrixGaussDiagLayerFF randmat randmat2 randvector tscalar multvector log_f Polygamma change_random_seed Psi BaseOptimizer Adam VMGNet loglike RandomStreams RandomState normal uniform normal concatenate sqrt uniform zeros astype floatX concatenate sum astype float32 zeros ravel range

Example implementation of the Bayesian neural network in: ***"Structured and Efficient Variational Deep Learning with Matrix Gaussian Posteriors"***, Christos Louizos & Max Welling, ICML 2016, ([https://arxiv.org/abs/1603.04733]()) This code is provided as is and will not be maintained / updated.

ANLGBOY/SoftFlow

['point cloud generation']

['SoftFlow: Probabilistic Framework for Normalizing Flow on Manifolds']

toy_example/lib/layers/diffeq_layers/__init__.py toy_example/generate2.py toy_example/lib/layers/diffeq_layers/basic.py pointclouds/metrics/pytorch_structural_losses/nn_distance.py toy_example/lib/layers/container.py toy_example/lib/layers/odefunc.py pointclouds/utils.py pointclouds/metrics/pytorch_structural_losses/setup.py pointclouds/args.py pointclouds/models/glow.py pointclouds/models/networks.py toy_example/lib/visualize_flow.py toy_example/generate1.py toy_example/lib/layers/__init__.py pointclouds/generate2.py pointclouds/evaluate.py toy_example/lib/layers/cnf.py toy_example/lib/layers/wrappers/cnf_regularization.py pointclouds/train.py toy_example/lib/layers/squeeze.py pointclouds/models/AF.py toy_example/train.py toy_example/lib/toy_data.py toy_example/lib/utils.py pointclouds/metrics/pytorch_structural_losses/__init__.py pointclouds/metrics/evaluation_metrics.py pointclouds/generate1.py pointclouds/metrics/pytorch_structural_losses/match_cost.py pointclouds/datasets.py get_parser get_args add_args ShapeNet15kPointClouds get_testset get_data_loaders get_trainset Uniform15KPC init_np_seed main get_test_loader evaluate_gen viz_save_recon get_viz_config viz_save_sample gen_samples main gen_samples viz_save_sample main get_viz_config main_worker get_init_data main standard_normal_logprob AverageValueMeter visualize_point_clouds validate_sample validate bernoulli_log_likelihood truncated_normal apply_random_rotation kl_diagnormal_diagnormal set_random_seed gaussian_log_likelihood validate_conditioned resume kl_diagnormal_stdnormal save reduce_tensor _pairwise_EMD_CD_ jensen_shannon_divergence _jsdiv EMD_CD lgan_mmd_cov knn compute_all_metrics jsd_between_point_cloud_sets distChamferCUDA unit_cube_grid_point_cloud entropy_of_occupancy_grid emd_approx distChamfer MatchCostFunction NNDistanceFunction remove ActNorm Flow Invertible1x1Conv AF ConcatSquashLinear remove fused_add_tanh_sigmoid_multiply ActNorm Invertible1x1Conv WN Glow SoftPointFlow Encoder get_transforms save_fig get_transforms save_fig generate get_transforms compute_loss inf_train_gen standard_normal_logprob build_model_tabular count_parameters RunningAverageMeter AverageMeter count_nfe inf_generator logsumexp isnan save_checkpoint count_total_time get_logger makedirs plt_flow plt_samples visualize_transform plt_potential_func plt_flow_samples plt_flow_density _flip CNF SequentialFlow Swish ODEfunc sample_rademacher_like Lambda ODEnet divergence_approx divergence_bf sample_gaussian_like SqueezeLayer squeeze unsqueeze ConcatSquashLinear RegularizedODEfunc add_argument ArgumentParser add_args parse_args join get_parser initial_seed seed print ShapeNet15kPointClouds print ShapeNet15kPointClouds DataLoader get_datasets DataLoader get_testset batch_size get_test_loader std_max list std_scale compute_all_metrics pprint test_std_n append cat item sample JSD flush items print write dumps numpy load join print SoftPointFlow multi_gpu_wrapper makedirs load_checkpoint eval load_state_dict set_initialized cuda open set_title set_xlim add_subplot axis draw close shape scatter set_zlim figure get_xlim get_ylim savefig range get_zlim set_ylim set_xlim add_subplot axis draw close shape scatter set_zlim figure savefig range set_ylim tr_max_sample_points std_max join format get_testset reconstruct viz_save_recon print std_scale get_trainset viz_save_sample DataLoader test_std_n sample range cuda gpu enumerate get_viz_config cates str set_title enumerate len squeeze stack append batch_size model SoftPointFlow multi_gpu_wrapper get_trainset DataLoader save cuda resume_checkpoint make_optimizer str all_points_mean std_max StepLR all_points_std get_testset view set_device DistributedSampler len random_rotate strftime std_scale epochs shape rank range SummaryWriter format init_process_group randn_like distributed resume set_initialized enumerate join int time collect apply_random_rotation print set_epoch std_min shuffle_idx train step array gpu add_scalar std_max view std_scale randn_like get_trainset shape DataLoader std_min iter next seed world_size get_args spawn warn set_random_seed device_count distributed main_worker get_init_data save_dir gpu clamp exp log clamp exp exp expand_as squeeze add_ copy_ shape normal_ all_reduce clone get_world_size size log seed manual_seed_all manual_seed set_title transpose set_xlim add_subplot axis draw close scatter get_ylim figure get_xlim set_zlim _renderer array get_zlim set_ylim bmm ones reshape print rand cos pi sin zeros to join reconstruct print EMD_CD batch_size iter save append numpy gpu cat update join append batch_size print size compute_all_metrics pprint iter save sample numpy JSD gpu cat load load_state_dict join list items makedirs eval item use_latent_flow add_scalar float match_cost bmm size transpose expand_as long list min mean distChamferCUDA distChamfer append emd_approx range cat list view size min contiguous expand distChamfer distChamferCUDA append emd_approx range cat update topk float size sqrt index_select to range cat to size min mean float update _pairwise_EMD_CD_ lgan_mmd_cov knn t ndarray reshape float32 float range reshape squeeze fit float warn unit_cube_grid_point_cloud unique zeros kneighbors len _jsdiv sum entropy warn sum ModuleList append remove_weight_norm sigmoid tanh subplot join invert_yaxis set_xlim close set_ticks scatter set_facecolor figure set_alpha dirname savefig set_ylim makedirs to numpy int64 save_fig inf_train_gen std_max data std_weight view batch_size model randn_like std_min to sum RandomState ones rand cos astype pi sqrt binomial vstack linspace sin append randint range getLogger addHandler StreamHandler info DEBUG setLevel INFO FileHandler __iter__ join save makedirs exp isinstance squeeze sum max pi AccNumEvals apply Accumulator apply tuple SequentialFlow map split set_title reshape invert_yaxis hstack pcolormesh set_ticks linspace meshgrid Tensor numpy prior_logdensity transform cmap set_title reshape invert_yaxis hstack set_xlim pcolormesh set_ticks set_facecolor linspace get_cmap meshgrid to set_ylim int set_title reshape hstack set_ticks imshow int64 cat linspace meshgrid inverse_transform to sum append split int transform set_title invert_yaxis set_xlim set_ticks int64 scatter append to numpy set_ylim split set_title invert_yaxis set_xlim set_ticks scatter set_ylim plt_flow subplot plt_samples plt_potential_func plt_flow_samples clf plt_flow_density size dim arange range sum size view contiguous size view contiguous

# SoftFlow: Probabilistic Framework for Normalizing Flow on Manifolds This repository provides the implementation of SoftFlow on toy dataset and point clouds. Move to each folder, follow the instructions and enjoy the results! ## Overview <p align="center"> <img src="assets/training_technique.png" height=256/> </p> Flow-based generative models are composed of invertible transformations between two random variables of the same dimension. Therefore, flow-based models cannot be adequately trained if the dimension of the data distribution does not match that of the underlying target distribution. In this paper, we propose SoftFlow, a probabilistic framework for training normalizing flows on manifolds. To sidestep the dimension mismatch problem, SoftFlow estimates a conditional distribution of the perturbed input data instead of learning the data distribution directly. We experimentally show that SoftFlow can capture the innate structure of the manifold data and generate high-quality samples unlike the conventional flow-based models. Furthermore, we apply the proposed framework to 3D point clouds to alleviate the difficulty of forming thin structures for flow-based models. The proposed model for 3D point clouds, namely SoftPointFlow, can estimate the distribution of various shapes more accurately and achieves state-of-the-art performance in point cloud generation. ## Results - Toy datasets <p align="center">

ANLGBOY/WaveNODE

['speech synthesis']

['WaveNODE: A Continuous Normalizing Flow for Speech Synthesis']

preprocessing.py train.py data.py odefunc.py hps.py args.py layers.py synthesize.py test_speed.py test_cll.py utils.py model.py parse_args _pad collate_fn_synthesize _pad_2d collate_fn LJspeechDataset Hyperparameters fused_add_tanh_sigmoid_multiply ActNorm CNF WaveNetPrior logabs MovingBatchNorm1d MBNLayer remove NODEBlock WaveNODE DELayer SqueezeLayer ActnormLayer ODEfunc fused_add_tanh_sigmoid_multiply_with_t WaveNet ODEnet divergence_approx sample_rademacher_like _process_utterance preprocess build_from_path write_metadata load_checkpoint build_model synthesize load_dataset load_checkpoint build_model evaluate load_dataset load_checkpoint build_model synthesize load_dataset synthesize evaluate build_model load_checkpoint save_checkpoint load_dataset train actnorm_init get_logger count_nfe mkdir add_argument ArgumentParser pad seed contiguous microsecond array append randint max range len max contiguous append array range len sigmoid tanh ModuleList append remove_weight_norm sum sigmoid tanh view ProcessPoolExecutor load join T astype float32 maximum pad log10 save max clip len build_from_path write_metadata makedirs print sum max data_path DataLoader LJspeechDataset print WaveNODE time format write_wav print synchronize size new_zeros Normal new_ones sample to numpy enumerate load join format print load_state_dict sum format model print write dumps eval flush enumerate write dumps flush write dumps flush model clip_grad_norm_ zero_grad format flush enumerate synth_interval log_interval time collect backward print write dumps parameters step array len time len array eval join format save state_dict print next iter AccNumEvals apply join format open n_layer_wvn str T norm temp batch_size n_channel_wvn split_period tol n_block tol_synth model_name scale scale_init d_i makedirs

# Pytorch Implementation of WaveNODE PyTorch implementation of WaveNODE ## Abstract In recent years, various flow-based generative models have been proposed to generate highfidelity waveforms in real-time. However, these models require either a well-trained teacher network or a number of flow steps making them memory-inefficient. In this paper, we propose a novel generative model called WaveNODE which exploits a continuous normalizing flow for speech synthesis. Unlike the conventional models,

APooladian/ProxLogBarrierAttack

['adversarial attack']

['A principled approach for generating adversarial images under non-smooth dissimilarity metrics']

mnist-example/model/utils.py proxlogbarrier_Top1.py mnist-example/model/__init__.py proxlogbarrier_Top5.py mnist-example/model/mnist.py InitMethods.py prox.py mnist-example/run_attack.py mnist-example/model/blocks.py simplex.py GaussianInitialize UniformInitialize BlurInitialize HeatSmoothing SafetyInitialize L2NormProx_Batch LinfNormProx L0NormProx_Batch L1NormProx Attack Top1Criterion Attack Top5Criterion L1BallProj _proj simplexproj indices SimplexProj prod Conv Linear LeNet View arange expand shape div_ zeros argmax list ones tolist expand append enumerate len

# ProxLogBarrierAttack Public repository for the ProxLogBarrier attack, described in [A principled approach for generating adversarial images under non-smooth dissimilarity metrics](https://arxiv.org/abs/1908.01667). Abstract: Deep neural networks perform well on real world data, but are prone to adversarial perturbations: small changes in the input easily lead to misclassification. In this work, we propose an attack methodology catered not only for cases where the perturbations are measured by Lp norms, but in fact any adversarial dissimilarity metric with a closed proximal form. This includes, but is not limited to, L1,L2,L_inf perturbations, and the L0 counting "norm", i.e. true sparseness. Our approach is a natural extension of a recent adversarial attack method, and eliminates the differentiability requirement of the metric. We demonstrate our algorithm, ProxLogBarrier, on the MNIST, CIFAR10, and ImageNet-1k datasets. We consider undefended and defended models, and show that our algorithm transfers to various datasets with little parameter tuning. Furthermore, we observe that ProxLogBarrier obtains the best results with respect to a host of modern adversarial attacks specialized for the L0 case. ## Implementation details Code is written in Python 3 and PyTorch 1.0. The implementation takes advantage of the GPU: a batch of images can be attacked at a given time. Hyperparameters are the defaults for the MNIST, CIFAR10, and ImageNet-1k datasets. We provide a pre-trained MNIST model (LeNet), with an example attack provided in `mnist-example/run_attack.py`. The attack is implemented as a class in `proxlogbarrier_Top1.py`. The Top5 version is included as well. ### Citation If you find the ProxLogBarrier attack useful in your scientific work, please cite as

ARMargolis/melanoma-pytorch

['network pruning']

['The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks']

model/data.py scripts/train_model.py model/model.py MelanomaDataset MelanomaNet BasicBlock BottleneckBlock

Goal: A PyTorch model for https://www.kaggle.com/c/siim-isic-melanoma-classification/data Technique: Pruning https://pytorch.org/tutorials/intermediate/pruning_tutorial.html#iterative-pruning Theory: Lottery Ticket Hypothesis https://arxiv.org/pdf/1803.03635.pdf # Setup ## Colab stuff To use the kaggle API in CoLab, follow the instructions here: [ https://medium.com/@move37timm/using-kaggle-api-for-google-colaboratory-d18645f93648#:~:text=To%20use%20the%20Kaggle%20API,of%20a%20file%20called%20'kaggle ] (some of the steps are present in the `setting_up_kaggle_credentials.ipynb` notebook, so you can copy them over) ## Dependencies We will use poetry to manage dependencies, so all you have to do is: 1. Clone the repo

ASMIftekhar/VSGNet

['human object interaction detection']

['VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions']

scripts_hico/HICO_eval/sample_complexity_analysis.py scripts/calculate_ap_classwise.py scripts_hico/train_test.py scripts/dataloader_vcoco.py scripts/prior_vcoco.py scripts_hico/helpers_preprocess.py scripts_hico/calculate_map_vcoco.py scripts_hico/model.py scripts_hico/HICO_eval/bbox_utils.py scripts/pred_vis.py scripts_hico/main.py scripts/model.py scripts_hico/pred_vis.py scripts/pool_pairing.py scripts/train_test.py scripts_hico/proper_inferance_file.py scripts/calculate_map_vcoco.py scripts_hico/HICO_eval/compute_map.py scripts/helpers_preprocess.py scripts/proper_inferance_file.py scripts_hico/calculate_ap_classwise.py scripts/main.py scripts_hico/pool_pairing.py setup.py scripts_hico/HICO_eval/load.py scripts_hico/HICO_eval/hico_constants.py scripts_hico/dataloader_hico.py vis_sub_obj_bboxes compute_area_batch join_bboxes_by_line add_bbox compute_iou compute_area compute_iou_batch vis_human_keypts vis_bboxes vis_bbox compute_ap eval_hoi compute_pr compute_normalized_pr load_gt_dets match_hoi ForkedPdb main HicoConstants load_pickle_object load_json_object read dump_json_object load_mat_object dump_pickle_object deserialize_object write load_yaml_object mkdir_if_not_exists dumps_json_object JsonSerializableClass serialize_object WritableToFile NumpyAwareJSONEncoder main compute_mAP polygon polygon_perimeter set_color min max compute_area zeros logical_and minimum compute_area_batch stack maximum min copy set_color polygon polygon_perimeter max range copy vis_bbox min copy line_aa max range circle vis_bboxes join_bboxes_by_line zip line_aa circle range copy compute_iou enumerate isnan any max arange cumsum array nan cumsum sum array nan int compute_ap format join compute_pr print match_hoi save append print join load_json_object append join load_json_object format starmap dump_json_object proc_dir print load_gt_dets num_processes mkdir_if_not_exists close set mkdir out_dir append parse_args Pool len decompress read loads compress write dumps encode compress write dumps dumps mkdir exists makedirs list items sorted compute_mAP HicoConstants bin_to_hoi_ids_json keys

# VSGNet ### [**VSGNet:Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions**](http://openaccess.thecvf.com/content_CVPR_2020/papers/Ulutan_VSGNet_Spatial_Attention_Network_for_Detecting_Human_Object_Interactions_Using_CVPR_2020_paper.pdf) [Oytun Ulutan*](https://sites.google.com/view/oytun-ulutan), [A S M Iftekhar*](https://sites.google.com/view/asmiftekhar/home), [B S Manjunath](https://vision.ece.ucsb.edu/people/bs-manjunath). Official repository of our [**CVPR 2020**](http://cvpr2020.thecvf.com/) paper.  ## Citing If you find this work useful, please consider our paper to cite: @InProceedings{Ulutan_2020_CVPR, author = {Ulutan, Oytun and Iftekhar, A S M and Manjunath, B. S.}, title = {VSGNet: Spatial Attention Network for Detecting Human Object Interactions Using Graph Convolutions},

AWLyrics/SeqGAN_Poem

['text generation']

['SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient']

lyric_preprocessing.py preprocessing.py bleu_calc.py dataloader.py mydis.py myG_beta.py rhyme.py seq_gan.py data/preprocess.py mygen.py id2poem.py session_save.py translate.py calc_bleu Input_Data_loader Gen_Data_loader Dis_dataloader load_data linear highway Discriminator Generator G_beta rhyme2 rhyme1 generate_samples pre_train_epoch_v2 generate_samples_v2 pre_train_epoch main translate_file translate write2file preprocess sentence_bleu SmoothingFunction zip pad_sequences fit_on_texts Tokenizer texts_to_sequences as_list Dis_dataloader num_batch create_batches Session open seed run str calc_bleu Generator generate_v2 G_beta g_update Discriminator sum range load_train_data close get_reward mean update_params ConfigProto Input_Data_loader time train_op print write pre_train_epoch_v2 generate_samples_v2 global_variables_initializer next_batch reset_pointer range extend generate generate_v2 extend next_batch reset_pointer range pretrain_step num_batch append next_batch reset_pointer range pretrain_step_v2 num_batch append next_batch reset_pointer range get int append

# SeqGAN Poem 修改 SeqGAN 用于诗歌生成（去除 Oracle LSTM） ## Pipeline - 收集唐诗预处理（preprocessing.py），唐诗[数据集](https://github.com/chinese-poetry/chinese-poetry)，选取了 5000 首五言律诗（20 词） - Tokenize，将中文字转换为 index（train.txt），建立词表（dict.pkl），记录 vocab size - Pretrain Generation - Adersarial Training ## Reference Paper: https://arxiv.org/abs/1609.05473 Original Repo: [SeqGAN](https://github.com/LantaoYu/SeqGAN)