这个函数onnx 12不支持 2021.01.07

Upsampling:上采样，被nn.functional.interpolate代替

        nn.Upsampling is deprecated. Use nn.functional.interpolate instead.

pytorch: MaxUnpool2d 与 Upsampling

  

   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      #-*-coding:utf-8-*-
     
    
   

    

     

    
    

     

      import torch
     
    
   

    

     

    
    

     

      import torch.nn as nn
     
    
   

    

     

    
    

     

      import numpy as np
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      # pool of non-square window
     
    
   

    

     

    
    

     

      ## In the simplest case, the output value of the layer
     
    
   

    

     

    
    

     

      # with input size (N,C,H,W)(N, C, H, W)(N,C,H,W),
     
    
   

    

     

    
    

     

      # output (N,C,Hout,Wout)(N, C, H_{out}, W_{out})(N,C,Hout​,Wout​) and kernel_size (kH,kW)(kH, kW)(kH,kW) can be precisely described as:
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      data = np.arange(64).reshape((8,8))
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      A = torch.Tensor(data.reshape(1,1,8,8))
     
    
   

    

     

    
    

     

      print('A=',A)
     
    
   

    

     

    
    

     

      ## MAX POOL
     
    
   

    

     

    
    

     

      maxpool = nn.MaxPool2d((2, 2), stride=(2, 2),return_indices=True)
     
    
   

    

     

    
    

     

      B,indices = maxpool(A)
     
    
   

    

     

    
    

     

      print('B=',B)
     
    
   

    

     

    
    

     
 
     
    
   

    

     

    
    

     

      ## Upsample
     
    
   

    

     

    
    

     

      Upsample = nn.Upsample(scale_factor=2, mode='bilinear
     
    
  
 

文章来源: blog.csdn.net，作者：网奇，版权归原作者所有，如需转载，请联系作者。

原文链接：blog.csdn.net/jacke121/article/details/112314799