search

LifeBeyondExpectations / ECCV22-PointMixer

[ECCV 2022] Official pytorch implementation of the paper, "PointMixer: MLP-Mixer for Point Cloud Understanding"
103 stars 12 forks source link

about class BilinearFeedForward(nn.Module) role in the model？ #14

Closed xiaobaitu123344 closed 2 years ago

xiaobaitu123344 commented 2 years ago

Hello author, I would like to ask you about the role of this module, I have collected a lot of information, read the principle of it, but do not understand its role in your model is what?

LifeBeyondExpectations commented 2 years ago

As stated in https://github.com/LifeBeyondExpectations/ECCV22-PointMixer/issues/14#issue-1465898006 Bilinear function is the one way of feature-wise aggregation using MLPs. However, instead of using MLPs, I directly call the function, torch.nn.Bilinear(...) for simplicity.

xiaobaitu123344 commented 2 years ago

Thank you for your detailed explanation.

LifeBeyondExpectations commented 2 years ago

if you have any question, please reopen the issue.