Coupled Attention Framework of Convolutional Neural Network Based on Computer Intelligence

Huoxiang Yang,Yongsheng Liang,Wei Liu,Fanyang Meng,Chao Li

DOI: https://doi.org/10.1155/2022/7859287

IF: 3.12

2022-08-06

Computational Intelligence and Neuroscience

Abstract:Using an attention mechanism based on the convolutional neural networks (CNNs) improves the performance of computer vision tasks by enhancing the representation of the features. The existing attention methods enhance the expression of the features by modeling the internal information of the features. However, due to the limited information flow of the previous features, these methods are difficult to calibrate features more completely. In this paper, we propose a Coupled Attention Framework (CAF) that is a simple attention framework for improving the performance of the existing attention methods. In the CAF, a coupling branch is added to an existing attention method to generate the input attention maps and enhance the input features of the convolution. The input attention is then spread to the output features through coupling between the input attention maps and convolution, the output features. The final result is the experimental results on various visual tasks. The results show that applying CAF to most of the existing attention methods can improve the performance with fewer parameters.

mathematical & computational biology,neurosciences

What problem does this paper attempt to address?

The paper was initially intended to propose a method called the Coupled Attention Framework (CAF), which is an improved attention mechanism framework based on Convolutional Neural Networks (CNN). However, the provided text is actually about the retraction notice of the paper, rather than the content of the paper itself. The retraction notice indicates that there was misconduct in the publication process of the paper, thus its reliability and integrity cannot be guaranteed. From the partial content of the original paper's abstract, the original goal of the study was to address the following issues: - Existing attention methods have difficulty fully calibrating features when modeling internal information of features because the information flow of previous features is limited. - A Coupled Attention Framework (CAF) was proposed, which improves the performance of existing attention methods by adding a coupled branch to generate input attention maps and enhance input features, then propagating this input attention information to output features. In short, the paper originally attempted to address the problem of how to more effectively extract and utilize feature information in computer vision tasks using attention mechanisms, particularly in spatial and channel dimensions. However, due to the retraction, the validity of these results and conclusions is no longer recognized.

Coupled Attention Framework of Convolutional Neural Network Based on Computer Intelligence

HAM: Hybrid Attention Module in Deep Convolutional Neural Networks for Image Classification

A Multi-scale Contextual Attention Mechanism for Convolutional Neural Networks

Cross-Layer Channel Attention Mechanism For Convolutional Neural Networks

Attention CoupleNet: Fully Convolutional Attention Coupling Network for Object Detection

An Attention Module for Convolutional Neural Networks

Multiscale Hybrid Convolutional Deep Neural Networks with Channel Attention

A Multi-Branch Feature Fusion Strategy Based on an Attention Mechanism for Remote Sensing Image Scene Classification

CAT: Learning to Collaborate Channel and Spatial Attention from Multi-Information Fusion

Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer

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

ICAFusion: Iterative cross-attention guided feature fusion for multispectral object detection

An Overview of the Attention Mechanisms in Computer Vision

Attentional Feature Fusion

BA-Net: Bridge Attention for Deep Convolutional Neural Networks

A Novel Multi-attention Fusion Convolution Neural Network for Blind Image Quality Assessment

Convolutional Attention Network for Electrical/Ultrasound Dual-Modal Fusion 2-D Image Reconstruction

Co-Saliency Detection With Co-Attention Fully Convolutional Network

Two-Branch Attention Network via Efficient Semantic Coupling for One-Shot Learning

A feature-wise attention module based on the difference with surrounding features for convolutional neural networks