What do you mean by replacing SFT layer by direct concatenation

[Jian-danai]

In table 1, you proposed the ablation study. Could you give the sample code for direct concatenation in the intermediate layers? I am not sure how to concatenate the kernel_code in the intermediate layers.

[Jian-danai]

Because after concatenation, the shape of the feature map will be HxWx(C+C'), so did you use convolution layer with kernel size 1*1, channel_in=C+C', channel_out=C, before feeding the concatenated feature map into the convolutions of the residual block?

Or by: using convolutions(channel_in=C+C', channel_out=C) in residual blocks? Then did you use relu before feeding the concatenated feature into convolutions and after the concatenation?