Back to Search Start Over

GateAttentionPose: Enhancing Pose Estimation with Agent Attention and Improved Gated Convolutions

Authors :
Feng, Liang
Shen, Zhixuan
Wen, Lihua
Li, Shiyao
Xu, Ming
Publication Year :
2024

Abstract

This paper introduces GateAttentionPose, an innovative approach that enhances the UniRepLKNet architecture for pose estimation tasks. We present two key contributions: the Agent Attention module and the Gate-Enhanced Feedforward Block (GEFB). The Agent Attention module replaces large kernel convolutions, significantly improving computational efficiency while preserving global context modeling. The GEFB augments feature extraction and processing capabilities, particularly in complex scenes. Extensive evaluations on COCO and MPII datasets demonstrate that GateAttentionPose outperforms existing state-of-the-art methods, including the original UniRepLKNet, achieving superior or comparable results with improved efficiency. Our approach offers a robust solution for pose estimation across diverse applications, including autonomous driving, human motion capture, and virtual reality.

Details

Database :
arXiv
Publication Type :
Report
Accession number :
edsarx.2409.07798
Document Type :
Working Paper