Something wrong in implementation of SCSE module

[John1231983]

Hello, thanks for sharing your implementation. I did not understand why do you add 1 in the return of SCSE module

LightNet/modules/misc.py

Line 110 in 04af225

return torch.add(chn_se, 1, spa_se)

I think it should be

return (chn_se + spa_se)

Basiclly, the ScSE did as follows: (1) convolution 1x1 x C for channel side and (2) convolution 1x1x1 for spatial size. Then add (1) and (2)

[ryohachiuma]

The author did not add 1 in the code.
If you want to add two tensors, you should use add module below.

torch.add(input, value=1, other, out=None)

It will calculate input + value * other.
(Ref: https://pytorch.org/docs/stable/torch.html)