search

DeLightCMU / RSC

This is the official implementation of Self-Challenging Improves Cross-Domain Generalization, ECCV2020
BSD 2-Clause "Simplified" License
160 stars 18 forks source link

Confusing naming conventions #9

Closed SirRob1997 closed 3 years ago

SirRob1997 commented 3 years ago

I find myself confused about the naming conventions/comments and you should probably be more precise here. If I understood it correctly, the channel_mean is the mean used in Channel-Wise RSC and computes the mean across the spatial domain while the spatial_mean is the mean used in Spatial-Wise RSC and computes the mean across the channel domain.

This is why we need F.interpolate in the Spatial-Wise RSC since it doesn't make sense for architectures with average pooling afterwards like mentioned here: #3. By the way: This isn't mentioned in the paper, what is the intuition behind that even?

Justinhzy commented 3 years ago

Hi, different convolutional layers should be able to locate discriminative spatial information consistently. The high-level convolutional layer tends to capture more semantic concepts while the low-level layer tends to capture more details. That's why spatial interpolation can be used in a convolutional network with average pooling.