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

TL;DR

ECA-Net introduces an efficient channel attention module that enhances CNN performance with minimal additional parameters by using local cross-channel interaction via adaptive 1D convolution, balancing accuracy and complexity.

Contribution

The paper proposes the ECA module, a novel attention mechanism that avoids dimensionality reduction, uses local cross-channel interaction, and adaptively selects convolution kernel size for efficiency.

Findings

ECA-Net improves Top-1 accuracy by over 2% on ImageNet with fewer parameters.

The ECA module reduces model complexity significantly compared to SENet.

Experimental results show ECA-Net outperforms existing attention modules in efficiency and effectiveness.

Abstract

Recently, channel attention mechanism has 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 for achieving better performance, which inevitably increase model complexity. To overcome the paradox of performance and complexity trade-off, this paper proposes an Efficient Channel Attention (ECA) module, which only involves a handful of parameters while bringing clear performance gain. By dissecting the channel attention module in SENet, we empirically show avoiding dimensionality reduction is important for learning channel attention, and appropriate cross-channel interaction can preserve performance while significantly decreasing model complexity. Therefore, we propose a local cross-channel interaction strategy without dimensionality reduction, which can be efficiently implemented via $1D$ convolution. Furthermore, we develop a method to adaptively select kernel size of $1D$ convolution, determining coverage of local cross-channel interaction. The proposed ECA module is efficient yet effective, e.g., the parameters and computations of our modules against backbone of ResNet50 are 80 vs. 24.37M and 4.7e-4 GFLOPs vs. 3.86 GFLOPs, respectively, and the performance boost is more than 2% in terms of Top-1 accuracy. We extensively evaluate our ECA module on image classification, object detection and instance segmentation with backbones of ResNets and MobileNetV2. The experimental results show our module is more efficient while performing favorably against its counterparts.