ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks
By Qilong Wang, Banggu Wu, Peng Fei Zhu, P. Li, Wangmeng Zuo, Qinghua Hu. 2019
6,251 citations recorded by the source corpus.
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 both efficient and 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.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Model Architectures, Efficiency & Compression
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Luc Van Gool
- Fahad Shahbaz Khan
- Jianfeng Gao
- Shiji Song
- Yu Qiao
- Jingdong Wang
- Gao Huang
- Salman Hameed Khan
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Ming-Hsuan Yang
- Dacheng Tao
- Xiaogang Wang
- Yi Yang
- Chen Change Loy
- Trevor Darrell
- Bernt Schiele
In Model Architectures
In Efficiency & Compression
Citations in the corpus
- Dynamic Neural Networks: A Survey (2021, background): “vide more statistics [125], or replacing FC layers with efficient 1D convolutions [126].”
- CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification (2021, methodology): “, SENet [18] uses channel-attention, CBAM [41] adds the spatial attention and ECANet [37] proposes an efficient channel attention to further improve SENet.”
- Large Separable Kernel Attention: Rethinking the Large Kernel Attention Design in CNN (2023, background): “Generally, they can be grouped into four categories: spatial attention [26], [28], [29], [31], channel attention [26], [27], [45], temporal attention [46], [47], and branch attention [48], [49].”
- Guided Attention Network for Object Detection and Counting on Drones (2019, methodology): “[41] improve the efficiency of SENet using a local cross-channel interaction strategy without dimension reduction.”
- LocalViT: Analyzing Locality in Vision Transformers (2021, positive): “In particular, we try ReLU6, h-swish [19], squeeze-and-excitation (SE) module [41], efficient channel attention (ECA) module [42], and their combinations.”
- Visual Attention Methods in Deep Learning: An In-Depth Survey (2022, methodology): “Efficient Channel Attention (ECA) [22] is based on squeeze”
- Efficient Multi-Scale Attention Module with Cross-Spatial Learning (2023, background): “channel reduction ratios yield better performance, it may bring side effect in extracting deep visual representations, which is explored the efficiency without dimensionality reduction in Efficient channel attention (ECA) [10].”
- AIM 2020 Challenge on Real Image Super-Resolution: Methods and Results (2020, methodology): “The spatial attention module is borrowed from EDVR [30] and the efficient attention module is borrowed from ECA-Net [29].”
- DaViT: Dual Attention Vision Transformers (2022, methodology): “6%) than both of the two variants using SE [28] and ECA [61] blocks, respectively.”
- Multiattention Network for Semantic Segmentation of Fine-Resolution Remote Sensing Images (2020, methodology): “The convolutional block attention module (CBAM) [59], selective kernel unit (SK unit) [60] and efficient channel attention module (ECA) [61] further boost the SE block’s performance.”
- Visual attention network (2022, background): “Besides, self-attention is a special attention that only considers the adaptability in spatial dimension but ignores the adaptability in channel dimension, which is also important for visual tasks [33,85,80,1].”
- SCSA: Exploring the Synergistic Effects Between Spatial and Channel Attention (2024, positive): “Existing plug-and-play attention methods can be primarily categorized into three types: channel attention [23, 27, 62, 55, 31, 41], spatial attention [22, 56, 42, 9], and hybrid channel-spatial attention [57, 40, 38, 14, 20, 65, 39, 61].”
- Compounding the Performance Improvements of Assembled Techniques in a Convolutional Neural Network (2020, methodology): “For example, we already are planning to reassemble various new studies such as AugMix [12] and ECA-net [36], which were recently published.”
- ICAFusion: Iterative Cross-Attention Guided Feature Fusion for Multispectral Object Detection (2023, methodology): “ECANet [26] proposes a local cross-channel interaction strategy with an adaptive one-dimensional convolution, which only involves a handful of parameters while bringing clear performance gain.”
- SegNeXt: Rethinking Convolutional Attention Design for Semantic Segmentation (2022, background): “Differently, the goal of using channel attention is to make the network selectively attend to those important objects, which has been demonstrated important in previous works [30, 9, 72].”
- AAU-Net: An Adaptive Attention U-Net for Breast Lesions Segmentation in Ultrasound Images (2022, background): “[42] Q. Wang, B. Wu, P. Zhu, P. Li, W. Zuo, and Q. Hu, “ECA-Net: Efficient channel attention for deep convolutional neural networks,” in Proc.”
- Attention-Based Adaptive Spectral–Spatial Kernel ResNet for Hyperspectral Image Classification (2020, methodology): “As shown in [55], the best weights of yi ∈ y (i = 1, .”