Huster-Hq
commented
1 year ago The bug has been fixed, because the start_label parameter in the slic is cancelled in some versions of skimage.segmentation.
Closed Huster-Hq closed 1 year ago
Huster-Hq
commented
1 year ago The bug has been fixed, because the start_label parameter in the slic is cancelled in some versions of skimage.segmentation.
File "/data1/Huqiang/polyp-segmentation/boxshrink-main/robust_boxshrink.py", line 41, in
res = generate_embedding_masks_for_dataset(dataset, TESTING_DIR, feature_extract_model, mean_f, mean_b, save_as_png=True)
File "/data1/Huqiang/polyp-segmentation/boxshrink-main/scripts/embeddings.py", line 405, in generate_embedding_masks_for_dataset
mean_background_embedding,
File "/data1/Huqiang/polyp-segmentation/boxshrink-main/scripts/embeddings.py", line 372, in get_embedding_mask_or_box
mean_background_embedding=mean_background_embedding,
File "/data1/Huqiang/polyp-segmentation/boxshrink-main/scripts/embeddings.py", line 283, in create_embedding_mask
model=model,
File "/data1/Huqiang/polyp-segmentation/boxshrink-main/scripts/embeddings.py", line 141, in get_foreground_background_embeddings
device=DEVICE,
File "/data1/Huqiang/polyp-segmentation/boxshrink-main/scripts/embeddings.py", line 55, in fill_embedding_matrix
smallest, largest = get_bbox_coordinates_one_box(all_superpixels_mask_tmp)
File "/data1/Huqiang/polyp-segmentation/boxshrink-main/scripts/embeddings.py", line 37, in get_bbox_coordinates_one_box
print(torch.max(all_x))
RuntimeError: max(): Expected reduction dim to be specified for input.numel() == 0. Specify the reduction dim with the 'dim' argument.
when superpixel = 0 , the all_superpixels_mask_tmp = 0, because of : all_superpixels_mask_tmp[all_superpixels_mask_tmp != superpixel] = 0 all_superpixels_mask_tmp[all_superpixels_mask_tmp > 0] = 1 can't get the (smallest_y, smallest_x)