Sizes of tensors must match except in dimension 1. Expected size 24 but got size 25 for tensor numbe

Posted 2022-04-12 AI浩

在做图像分割的时候遇到了错误，错误如下：

File "D:/segmentation/Pytorch-UNet-master/train.py", line 193, in <module>
    amp=args.amp)
  File "D:/segmentation/Pytorch-UNet-master/train.py", line 88, in train_net
    masks_pred = net(images)
  File "D:\\ProgramData\\Anaconda3\\envs\\mytorch\\lib\\site-packages\\torch\\nn\\modules\\module.py", line 1110, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\\ProgramData\\Anaconda3\\envs\\mytorch\\lib\\site-packages\\segmentation_models_pytorch\\base\\model.py", line 16, in forward
    decoder_output = self.decoder(*features)
  File "D:\\ProgramData\\Anaconda3\\envs\\mytorch\\lib\\site-packages\\torch\\nn\\modules\\module.py", line 1110, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\\ProgramData\\Anaconda3\\envs\\mytorch\\lib\\site-packages\\segmentation_models_pytorch\\unet\\decoder.py", line 121, in forward
    x = decoder_block(x, skip)
  File "D:\\ProgramData\\Anaconda3\\envs\\mytorch\\lib\\site-packages\\torch\\nn\\modules\\module.py", line 1110, in _call_impl
    return forward_call(*input, **kwargs)
  File "D:\\ProgramData\\Anaconda3\\envs\\mytorch\\lib\\site-packages\\segmentation_models_pytorch\\unet\\decoder.py", line 39, in forward
    x = torch.cat([x, skip], dim=1)
RuntimeError: Sizes of tensors must match except in dimension 1. Expected size 24 but got size 25 for tensor number 1 in the list.

从错误的分析得知是维度没有对上。将x和skip的维度打印出来：

torch.Size([2, 448, 24, 18]) torch.Size([2, 160, 25, 18])

这就对应上上面的错误了。
出现这个错误的原因是在输入图片的时候，没有将图像resize成512×512大小，导致维度不一致！
将图片resize后就可以解决。
在打印x和skip的维度：

torch.Size([2, 448, 32, 32]) torch.Size([2, 160, 32, 32])