What is the difference between AutoCrop layer used here and traditional Crop layer used in HED and FCNs

yun-liu / RCF

Richer Convolutional Features for Edge Detection

Other

753 stars 259 forks source link

Open xuyuting45 opened 6 years ago

frankie-yanfeng commented 5 years ago

I guess the main difference is no manual offset(the middle) calculation in the layer parameters.

yun-liu commented 5 years ago

You are right! And sorry for missing this question!

qinyi20060410 commented 5 years ago

You are right! And sorry for missing this question!

那么如果在RCF的代码的基础上进行HED的重新训练，可以理解为在train的prototxt中用AutoCrop代替Crop吗？

yun-liu commented 5 years ago

@qinyi20060410 不太理解你的意思。RCF和HED的不同主要是网络不一样，不是Crop