p2oileen commited on
Commit
1ceb75c
1 Parent(s): bc9223a

add README

Browse files
Files changed (3) hide show
  1. .gitignore +0 -1
  2. MAIN.png +0 -0
  3. README.md +104 -1
.gitignore CHANGED
@@ -7,6 +7,5 @@ x264/
7
  *.mp3
8
  *.mp4
9
  filelist.txt
10
- *.png
11
  complex_infer.sh
12
  __pycache__/
 
7
  *.mp3
8
  *.mp4
9
  filelist.txt
 
10
  complex_infer.sh
11
  __pycache__/
MAIN.png ADDED
README.md CHANGED
@@ -1 +1,104 @@
1
- # CoNR
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # CoNR
2
+ # CoNR: Collaborative Neural Rendering using Anime Character Sheets
3
+
4
+ ---
5
+
6
+ ##[HomePage](https://conr.ml) | Colab | [arXiv](https://arxiv.org/abs/2207.05378)
7
+
8
+ ![image](MAIN.png)
9
+
10
+
11
+ ## Introduction
12
+
13
+ This project is the official implement of [Collaborative Neural Rendering using Anime Character Sheets](https://arxiv.org/abs/2207.05378), which aims to genarate vivid dancing videos from hand-drawn anime character sheets(ACS). Watch more demos in our [HomePage](https://conr.ml).
14
+
15
+ ## Usage
16
+
17
+ #### Prerequisites
18
+
19
+ * NVIDIA GPU + CUDA + CUDNN
20
+ * Python 3.6
21
+
22
+ #### Installation
23
+
24
+ * Clone this repository
25
+
26
+ ```bash
27
+ git clone https://github.com/megvii-research/CoNR
28
+ ```
29
+
30
+ * Dependencies
31
+
32
+ To install all the dependencies, please run the following commands.
33
+
34
+ ```bash
35
+ cd CoNR
36
+ pip install -r requirements.txt
37
+ ```
38
+
39
+ * Download Weights
40
+ Download weights from Google Drive. Alternatively, you can download from [Baidu Netdisk](https://pan.baidu.com/s/1U11iIk-DiJodgCveSzB6ig?pwd=RDxc) (password:RDxc).
41
+
42
+ ```
43
+ mkdir weights && cd weights
44
+ wget https://drive.google.com/file/d/1M1LEpx70tJ72AIV2TQKr6NE_7mJ7tLYx/view?usp=sharing
45
+ wget https://drive.google.com/file/d/1YvZy3NHkJ6gC3pq_j8agcbEJymHCwJy0/view?usp=sharing
46
+ wget https://drive.google.com/file/d/1AOWZxBvTo9nUf2_9Y7Xe27ZFQuPrnx9i/view?usp=sharing
47
+ wget https://drive.google.com/file/d/19jM1-GcqgGoE1bjmQycQw_vqD9C5e-Jm/view?usp=sharing
48
+ ```
49
+
50
+ #### Prepare inputs
51
+ We prepared two Ultra-Dense Pose sequences for two characters, you can generate more UDPs via 3D models and motions.
52
+ [Baidu Netdisk](https://pan.baidu.com/s/1hWvz4iQXnVTaTSb6vu1NBg?pwd=RDxc) (password:RDxc)
53
+
54
+ ```
55
+ # for short hair girl
56
+ wget https://drive.google.com/file/d/11HMSaEkN__QiAZSnCuaM6GI143xo62KO/view?usp=sharing
57
+ unzip short_hair.zip
58
+ mv short_hair/ poses/
59
+
60
+ # for double ponytail girl
61
+ wget https://drive.google.com/file/d/1WNnGVuU0ZLyEn04HzRKzITXqib1wwM4Q/view?usp=sharing
62
+ unzip double_ponytail.zip
63
+ mv double_ponytail/ poses/
64
+ ```
65
+
66
+ We provide sample inputs of anime character sheets, you can also draw more by yourself.
67
+
68
+ ```
69
+ # for short hair girl
70
+ wget https://drive.google.com/file/d/1r-3hUlENSWj81ve2IUPkRKNB81o9WrwT/view?usp=sharing
71
+ unzip short_hair_images.zip
72
+ mv short_hair_images/ character_sheet/
73
+
74
+ # for double ponytail girl
75
+ wget https://drive.google.com/file/d/1XMrJf9Lk_dWgXyTJhbEK2LZIXL9G3MWc/view?usp=sharing
76
+ unzip double_ponytail_images.zip
77
+ mv double_ponytail_images/ character_sheet/
78
+ ```
79
+
80
+ #### RUN!
81
+ We provide two ways: with web UI or via terminal.
82
+
83
+ * with web UI (powered by [Streamlit](https://streamlit.io/))
84
+
85
+ ```
86
+ streamlit run streamlit.py --server_port=8501
87
+ ```
88
+ then open your browser and visit `localhost:8501`, follow the instructions to genarate video.
89
+
90
+ * via terminal
91
+
92
+ ```
93
+ mkdir {dir_to_save_result}
94
+
95
+ python3 -m torch.distributed.launch \
96
+ --nproc_per_node=1 train.py --mode=test \
97
+ --world_size=1 --dataloaders=2 \
98
+ --test_input_poses_images={dir_to_poses} \
99
+ --test_input_person_images={dir_to_character_sheet} \
100
+ --test_output_dir={dir_to_save_result} \
101
+ --test_checkpoint_dir={dir_to_weights}
102
+
103
+ ffmpeg -r 30 -y -i {dir_to_save_result}/%d.png -r 30 -c:v libx264 output.mp4 -r 30
104
+ ```