如何在Keras中的卷积层之间添加跳过连接

Question

我想在keras中的残留块之间添加跳过连接。这是我目前的实现，因为张量具有不同的形状。

函数看起来像这样：

def build_res_blocks(net, x_in, num_res_blocks, res_block, num_filters, res_block_expansion, kernel_size, scaling):
    net_next_in = net
    for i in range(num_res_blocks):
        net = res_block(net_next_in, num_filters, res_block_expansion, kernel_size, scaling)
    
        # net tensor shape: (None, None, 32)
        # x_in tensor shape: (None, None, 3)
        # Error here, net_next_in should be in the shape of (None, None, 32) to be fed into next layer
        net_next_in = Add()([net, x_in]) 
    return net

但我得到

ValueError: Operands could not be broadcast together with shapes (None, None, 32) (None, None, 3)

如何将这些张量添加或合并到正确的形状中（无，无，32）？如果这不是正确的方法，您如何才能达到预期的结果？

这就是res_block的样子：

def res_block(x_in, num_filters, expansion, kernel_size, scaling):
    x = Conv2D(num_filters * expansion, kernel_size, padding='same')(x_in)
    x = Activation('relu')(x)
    x = Conv2D(num_filters, kernel_size, padding='same')(x)
    x = Add()([x_in, x])
return x

Answer 1

您不能添加不同形状的张量。您可以将它们与keras.layers.concatenate相连，但这会让您张开形状[None, None, 35]。

另外，请看一下RESNET50在Keras实施。它们的残留块在快捷方式中具有1x1xc卷积，对于要添加尺寸的情况不同。