Multi-Grained Feature Pruning for Video-Based Human Pose Estimation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Zhigang, Fan, Shaojing, Liu, Zhenguang, Wu, Zheqi, Wu, Sifan, Jiao, Yingying
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916646398984192
author Wang, Zhigang
Fan, Shaojing
Liu, Zhenguang
Wu, Zheqi
Wu, Sifan
Jiao, Yingying
author_facet Wang, Zhigang
Fan, Shaojing
Liu, Zhenguang
Wu, Zheqi
Wu, Sifan
Jiao, Yingying
contents Human pose estimation, with its broad applications in action recognition and motion capture, has experienced significant advancements. However, current Transformer-based methods for video pose estimation often face challenges in managing redundant temporal information and achieving fine-grained perception because they only focus on processing low-resolution features. To address these challenges, we propose a novel multi-scale resolution framework that encodes spatio-temporal representations at varying granularities and executes fine-grained perception compensation. Furthermore, we employ a density peaks clustering method to dynamically identify and prioritize tokens that offer important semantic information. This strategy effectively prunes redundant feature tokens, especially those arising from multi-frame features, thereby optimizing computational efficiency without sacrificing semantic richness. Empirically, it sets new benchmarks for both performance and efficiency on three large-scale datasets. Our method achieves a 93.8% improvement in inference speed compared to the baseline, while also enhancing pose estimation accuracy, reaching 87.4 mAP on the PoseTrack2017 dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2503_05365
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Multi-Grained Feature Pruning for Video-Based Human Pose Estimation
Wang, Zhigang
Fan, Shaojing
Liu, Zhenguang
Wu, Zheqi
Wu, Sifan
Jiao, Yingying
Computer Vision and Pattern Recognition
Human pose estimation, with its broad applications in action recognition and motion capture, has experienced significant advancements. However, current Transformer-based methods for video pose estimation often face challenges in managing redundant temporal information and achieving fine-grained perception because they only focus on processing low-resolution features. To address these challenges, we propose a novel multi-scale resolution framework that encodes spatio-temporal representations at varying granularities and executes fine-grained perception compensation. Furthermore, we employ a density peaks clustering method to dynamically identify and prioritize tokens that offer important semantic information. This strategy effectively prunes redundant feature tokens, especially those arising from multi-frame features, thereby optimizing computational efficiency without sacrificing semantic richness. Empirically, it sets new benchmarks for both performance and efficiency on three large-scale datasets. Our method achieves a 93.8% improvement in inference speed compared to the baseline, while also enhancing pose estimation accuracy, reaching 87.4 mAP on the PoseTrack2017 dataset.
title Multi-Grained Feature Pruning for Video-Based Human Pose Estimation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.05365