细粒度分类：MC-Loss论文笔记——The Devil is in the Channels: Mutual-Channel Loss for FGVC

综述

论文题目：《The Devil is in the Channels: Mutual-Channel Loss for Fine-Grained Image Classification》
期刊与时间：IEEE Transactions on Image Processing 2020 (TIP 2020)

论文地址：https://arxiv.org/pdf/2002.04264

源码地址(PyTorch版本)：https://github.com/dongliangchang/Mutual-Channel-Loss

针对领域：细粒度图像分类(FGVC)

源码笔记：https://blog.csdn.net/qq_50001789/article/details/131353832?spm=1001.2014.3001.5501

主要思想

  细粒度分类任务中，由于类别间的视觉差异往往很微小，并且通常存在于局部细节中，因此