FasterNet

FasterNet copied to clipboard

PConv is similar to GhostConv

1

Dear author, in order to use redundant feature information, PConv is proposed in this paper. However, Partial convolution is so similar to Ghost convolution. Ghost convolution uses half of the...

baowenshuwh

Does using GELU or RELU have a critical impact on the performance of the T0 model?

1

I noticed that you use GELU in small models like T0 and T1 and RELU in larger models like T2, is this intentional or just an oversight?

MenSanYan

some problem

4

I follow your idea to train my model using dataset about face， but results are comparable to resnet（params are comparable）， what do you think may be the problem？Thank you .

luckydog11

Hello, I have read your paper and found it truly impressive. However, I have a small question: when using only 1/4 of the channels for computation, how can we ensure...

Zhangyuhaoo