0%

『论文笔记』Squeeze-and-Excitation Networks

基本信息

1
2
3
4
5
6
7
@inproceedings{hu2018squeeze,
title={Squeeze-and-excitation networks},
author={Jie Hu and Li Shen and Gang Sun},
booktitle={IEEE Conference on Computer Vision and Pattern Recognition},
pages={7132--7141},
year={2018}
}

这篇文章主要是想看看实验细节,看看我的工作能不能用在该结构上面。

摘要

提出了一个 SE 模块(Squeezeand-Excitation block)用于通过对通道相关性建模以自适应调整通道特征响应,该结构在 SOTA 模型中以最小的代价实现了显著的性能提升。该方法在 ILSVRC 2017 分类比赛中获得了第一名。

介绍

简单介绍背景和一些相关的研究进展

方法

SE-pipeline

SE-ResNet-module

SE-Inception-module

主要看看文章的思路和逻辑

具体细节

包括公式推导和实验部分的一些参数细节等等,不是很重要的文章可以不看

实验部分

Pytorch 实现如下

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
class SELayer(nn.Module):
def __init__(self, channel, reduction=16):
super(SELayer, self).__init__()
self.avg_pool = nn.AdaptiveAvgPool2d(1)
self.fc = nn.Sequential(
nn.Linear(channel, channel // reduction, bias=False),
nn.ReLU(inplace=True),
nn.Linear(channel // reduction, channel, bias=False),
nn.Sigmoid()
)

def forward(self, x):
b, c, _, _ = x.size()
y = self.avg_pool(x).view(b, c)
y = self.fc(y).view(b, c, 1, 1)
return x * y.expand_as(x)

结论

该论文提出了 SE 模块结构,提高了网络的表达能力,通过动态通道特征调整。给出了以往结构在建模通道特征依赖的一些局限。该结构对相关领域如网络剪枝可能会有帮助。

其他

一些杂项

参考资料

--- ♥ end ♥ ---

欢迎关注我呀~