Caffe Smooth_L1_Loss_Layer 问答

最新推荐文章于 2024-01-23 17:30:55 发布

maybepossible

最新推荐文章于 2024-01-23 17:30:55 发布

阅读量2.3k

点赞数 1

分类专栏： Machine Learning 文章标签： caffe 深度学习

本文链接：https://blog.csdn.net/WL2002200/article/details/53994860

版权

Machine Learning 专栏收录该内容

28 篇文章 1 订阅

订阅专栏

问：参数中设置sigma原因是什么？

rbg答：As sigma -> inf the loss approaches L1 (abs) loss. Setting sigma = 3, makes the transition point from quadratic to linear happen at |x| <= 1 / 3**2 (closer to the origin). The reason for doing this is because the RPN bbox regression targets are not normalized by their stdev (unlike in Fast R-CNN), because the statistics of the targets are changing constantly throughout learning. In a future update I may simply replace smooth L1 with (hard) L1 which I believe will likely work as well and be simpler (no sigma, etc.).

问：为什么Smooth_L1_loss对target box(bottom[1])也进行反向传播？

rbg答：Smooth L1 loss can be used in cases where you do want to bprop to both inputs (e.g., in a “siamese” network). In the case of Fast R-CNN, we don’t need derivatives for the bbox regression labels, but the layer is more general than its use in Fast R-CNN.