ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks

Qilong WangBanggu WuPengfei ZhuPeihua LiWangmeng ZuoQinghua Hu

   Papers with code   Abstract  PDF

Channel attention has recently demonstrated to offer great potential in improving the performance of deep convolutional neural networks (CNNs). However, most existing methods dedicate to developing more sophisticated attention modules to achieve better performance, inevitably increasing the computational burden... (read more)

Benchmarked Models

RANK
MODEL
REPO
CODE RESULT
PAPER RESULT
ε-REPRODUCED
BUILD
4
ECA-Net152
78.9%
78.9%
5
ECA-Net101
78.6%
78.6%
6
ECA-Net50
77.4%
77.4%
7
ECA-Net34
74.2%
74.2%
8
ECA-MobileNet_v2
72.6%
72.6%
9
ECA-Net18
70.9%
70.9%