Spaces:
Build error
Build error
| import gradio as gr | |
| from huggingface_hub import snapshot_download | |
| import tensorflow as tf | |
| from PIL import Image | |
| import numpy as np | |
| from object_detection.utils import label_map_util | |
| from object_detection.utils import visualization_utils as viz_utils | |
| # Path to the label map | |
| PATH_TO_LABELS = 'data/label_map.pbtxt' | |
| category_index = label_map_util.create_category_index_from_labelmap(PATH_TO_LABELS, use_display_name=True) | |
| def pil_image_as_numpy_array(pilimg): | |
| img_array = tf.keras.utils.img_to_array(pilimg) | |
| img_array = np.expand_dims(img_array, axis=0) | |
| return img_array | |
| def load_model(repo_id): | |
| download_dir = snapshot_download(repo_id) | |
| saved_model_dir = os.path.join(download_dir, "saved_model") | |
| detection_model = tf.saved_model.load(saved_model_dir) | |
| return detection_model | |
| def predict(pilimg): | |
| image_np = pil_image_as_numpy_array(pilimg) | |
| return predict_combined_models(image_np, detection_model1, detection_model2) | |
| def predict_combined_models(image_np, model1, model2): | |
| # Process with first model | |
| results1 = model1(image_np) | |
| result1 = {key:value.numpy() for key,value in results1.items()} | |
| # Process with second model | |
| results2 = model2(image_np) | |
| result2 = {key:value.numpy() for key,value in results2.items()} | |
| # Visualization for model 1 | |
| image_np_with_detections = image_np.copy() | |
| viz_utils.visualize_boxes_and_labels_on_image_array( | |
| image_np_with_detections[0], | |
| result1['detection_boxes'][0], | |
| (result1['detection_classes'][0]).astype(int), | |
| result1['detection_scores'][0], | |
| category_index, | |
| use_normalized_coordinates=True, | |
| max_boxes_to_draw=200, | |
| min_score_thresh=.60, | |
| agnostic_mode=False, | |
| line_thickness=2) | |
| # Visualization for model 2 (can adjust styles to differentiate) | |
| viz_utils.visualize_boxes_and_labels_on_image_array( | |
| image_np_with_detections[0], | |
| result2['detection_boxes'][0], | |
| (result2['detection_classes'][0]).astype(int), | |
| result2['detection_scores'][0], | |
| category_index, | |
| use_normalized_coordinates=True, | |
| max_boxes_to_draw=200, | |
| min_score_thresh=.60, | |
| agnostic_mode=False, | |
| line_thickness=2) | |
| # Combine and return final image | |
| result_pil_img = tf.keras.utils.array_to_img(image_np_with_detections[0]) | |
| return result_pil_img | |
| # Load your models | |
| REPO_ID1 = "dtyago/23a052w-iti107-assn2_tfodmodel" | |
| REPO_ID2 = "dtyago/23a052w-iti107-assn2_tfodmodel_run1" | |
| detection_model1 = load_model(REPO_ID1) | |
| detection_model2 = load_model(REPO_ID2) | |
| # Gradio interface | |
| gr.Interface( | |
| fn=predict, | |
| inputs=gr.Image(type="pil"), | |
| outputs=gr.Image(type="pil") | |
| ).launch(share=True) | |