RuntimeError: output with shape [1, 256, 192] doesn't match the broadcast shape [3, 256, 192]

[steven5clu884]

I clone the repo, and then create an anaconda environment
I then type in python test.py
I have checkpoints
── GMM
│   └── gmm_final.pth
└── TOM
└── tom_final.pth

My traceback looks like
Traceback (most recent call last):
File "test.py", line 229, in
main()
File "test.py", line 215, in main
test_gmm(opt, test_loader, model, board)
File "test.py", line 86, in test_gmm
for step, inputs in enumerate(test_loader.data_loader):
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 521, in next
data = self._next_data()
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1203, in _next_data
return self._process_data(data)
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1229, in _process_data
normalize1: torch.Size([3, 256, 192])
data.reraise()
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/_utils.py", line 434, in reraise
raise exception
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 49, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/utils/data/utils/fetch.py", line 49, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/wizzerking/develop/cp-vton-plus/cp_dataset.py", line 150, in getitem
shape_ori = self.transform(parse_shape_ori) # [-1,1]
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torchvision/transforms/transforms.py", line 61, in call
img = t(img)
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1102, in call_impl
return forward_call(*input, **kwargs)
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torchvision/transforms/transforms.py", line 226, in forward
return F.normalize(tensor, self.mean, self.std, self.inplace)
File "/home/wizzerking/develop/anaconda3/envs/cpvton/lib/python3.7/site-packages/torchvision/transforms/functional.py", line 352, in normalize
print("normalize1:",tensor.sub(mean).div(std).shape)
RuntimeError: output with shape [1, 256, 192] doesn't match the broadcast shape [3, 256, 192]

Things I have tried
in cp_dataset.py add .convert('RGB')
after each call top Image.open so for instance

   if self.stage == 'GMM':
        c = Image.open(osp.join(self.data_path, 'cloth', c_name))
        c = c.convert('RGB')
    
        cm = Image.open(osp.join(self.data_path, 'cloth-mask', c_name)).convert('L')
        cm = cm.convert('RGB')

This does not help; also adding to the person image

person image

    im = Image.open(osp.join(self.data_path, 'image', im_name))
    im = im.convert('RGB')

does not help

[rocketeerli]

you need to change you pytorch to 0.4.
Or you can just do as follows:
change the CPDataset class in cp_dataset.py
self.transform = transforms.Compose([ \ transforms.ToTensor(), \ transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])
to
self.transform = transforms.Compose([ \ transforms.ToTensor(), \ transforms.Normalize((0.5,), (0.5,))])

[slaifan]

Hi,

I am having the same problem, I couldn't find torch==0.4. I believe it is deprecated (please correct me if wrong).
Changing the Normalization code didn't work either as I am still getting the same error

Thanks in advance!

[zakmicallef]

If you updated your torch .view needs to be replaced with .reshape on network.py on line 135.

[niranjanakella]

In cp_dataset.py, line 30

Replace:

self.transform = transforms.Compose([
             transforms.ToTensor(),
             transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])

With this line:

self.transform = transforms.Compose([
            transforms.ToTensor(),
            transforms.Normalize((0.5,), (0.5,))])

And also in network.py, line 35
Replace:

x = x.view(x.size(0), -1)

With this line:

x = x.reshape(x.size(0), -1)