损失函数Target size must be the same as input size

拼搏现实的明天。 2022-02-25 08:57 1319阅读 0赞

## 出错程序 ##

criterion_modality = torch.nn.BCEWithLogitsLoss()
    
    label = Variable(label.cuda())
    
    loss = criterion_modality(outRGB, label) # 出错行

## 问题 ##

File "C:\Users\Rain\AppData\Local\Programs\Python\Anaconda.3.5.1\envs\python35\python35\lib\site-packages\torch\nn\modules\module.py", line 491, in __call__
        result = self.forward(*input, **kwargs)
      File "C:\Users\Rain\AppData\Local\Programs\Python\Anaconda.3.5.1\envs\python35\python35\lib\site-packages\torch\nn\modules\loss.py", line 500, in forward
        reduce=self.reduce)
    
      File "C:\Users\Rain\AppData\Local\Programs\Python\Anaconda.3.5.1\envs\python35\python35\lib\site-packages\torch\nn\functional.py", line 1514, in binary_cross_entropy_with_logits
    
        raise ValueError("Target size ({}) must be the same as input size ({})".format(target.size(), input.size()))
    
    ValueError: Target size (torch.Size([32])) must be the same as input size (torch.Size([32, 2]))

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L1dKX01laU1laQ_size_16_color_FFFFFF_t_70]

## 原因 ##

input 和 target 尺寸不匹配，参见官网  
![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L1dKX01laU1laQ_size_16_color_FFFFFF_t_70 1]

## 解决办法 ##

1.  换成支持输入input 和 输出output尺度不匹配的损失函数，torch只有如下损失函数符合要求  
    [交叉熵损失函数][Link 1] （[推导过程][Link 2]）–[样例][Link 3]  
    负对数似然损失函数  
    评价相似度的损失  
    三元组损失函数  
    ![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L1dKX01laU1laQ_size_16_color_FFFFFF_t_70 2]
2.  暴力解法，直接 reshape(input) 改变数组的形状，强制 input 形状为 output 的形状，不要觉得仿佛丢失了什么信息量，但实际没有丢失关键特征，不会影响实验效果，前提是reshape参数调好

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L1dKX01laU1laQ_size_16_color_FFFFFF_t_70]: /images/20220225/8f639f23bddd4df68c70972707d26e6a.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L1dKX01laU1laQ_size_16_color_FFFFFF_t_70 1]: /images/20220225/73a5af541d1a44c6859a33ec4c268be0.png
[Link 1]: https://pytorch.org/docs/stable/_modules/torch/nn/modules/loss.html#CrossEntropyLoss
[Link 2]: https://www.jianshu.com/p/b07f4cd32ba6
[Link 3]: https://zhuanlan.zhihu.com/p/35709485
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L1dKX01laU1laQ_size_16_color_FFFFFF_t_70 2]: /images/20220225/bdfeb602855141aa8fa324fb0bc4d616.png