yeates / mmt Goto Github PK

Hello, thank you for your excellent work! I used utils/generate_face_parsing.py to generate the segmentation maps for CelebA-HQ, but nothing was generated.

The edge_loss and seg_loss are increasing accumulatively when I train the OST dataset, whether it is the code 'weight = (label_anno.view(bs, -1) == (self.n_cate_anno-1)).sum(1) == h * w # for cityscapes datasets segmentation' to cause the error in training OST dataset.

and the training sample images are wrong, as the predicted segmentation map, predicted edge map and the inpainted image are error. And this wrong sample image usually occurs after 20000 training iters.

Traceback (most recent call last):
File "D:/研究生/图像修复/文献代码/Unbiased Multi-Modality Guidance for Image Inpainting/MMT-main/train.py", line 53, in
run()
File "D:/研究生/图像修复/文献代码/Unbiased Multi-Modality Guidance for Image Inpainting/MMT-main/train.py", line 40, in run
dataloader = DataLoader(dataset, batch_size = args.batch_size, shuffle = True, num_workers = args.n_threads)
File "D:\Tools\Anaconda\envs\magic\lib\site-packages\torch\utils\data\dataloader.py", line 262, in init
sampler = RandomSampler(dataset, generator=generator) # type: ignore
File "D:\Tools\Anaconda\envs\magic\lib\site-packages\torch\utils\data\sampler.py", line 103, in init
raise ValueError("num_samples should be a positive integer "
ValueError: num_samples should be a positive integer value, but got num_samples=0

Process finished with exit code 1

I tried your model and this is the result I got back. The left side is the predicted area. It seems to fail on skin tone, hair, ear and jaw line. Otherwise it's pretty good. Could you post your results so I have something to compare this to?

Thanks for this amazing work but I am unable to find pretrained weights for OST and Cityscape dataset. Can you please provide those.