【pytorch】torch.nn.AdaptiveMaxPool2d 自适应池化

https://pytorch.org/docs/master/generated/torch.nn.AdaptiveMaxPool2d.html#torch.nn.AdaptiveMaxPool2d

 

torch.nn.AdaptiveMaxPool2d

• output_size – the target output size of the image of the form H x W. Can be a tuple (H, W) or a single H for a square image H x H. H and W can be either a int, or None which means the size will be the same as that of the input.

• return_indices – if True, will return the indices along with the outputs. Useful to pass to nn.MaxUnpool2d. Default: False

GAP的优势在于：

各个类别于Feature Map 之间的联系更加直观（相比与全连接层的黑箱来说），Feature Map 被转化为分类概率也更加容易，因为在GAP中没有参数需要调，所以避免了过拟合的问题。

GAP汇总了空间信息