GENet/ESPNet

• GENet(更泛化的SEnet,有带参数和不参数的模块)

原文：https://blog.csdn.net/dgyuanshaofeng/article/details/84179196

SENet之后，Jie Hu和Li Shen等人又提出GENet[1]。单位：Momenta, Visual Geometry Group University of Oxford  ；

[1] Gather-Excite: Exploiting Feature Context in Convolutional Neural Networks NIPS 2018 [paper][code]

 

• GCNet：当Non-local遇见SENet

• 模块设计之 SKNet, GCNet, GloRe, Octave

 

• ESPNet/V2 (v1针对分割任务,就是先用pointwise卷积，然后用金字塔的空洞卷积，HFF(Hier feature fusion)，v2更泛化的模块->针对mobileNet and shufflenet)

原文链接：https://arxiv.org/abs/1803.06815v2

相关repo：https://github.com/sacmehta/ESPNet

论文学习笔记（四）ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation