Foundation Model for Skeleton-Based Human Action Understanding

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Hongsong, Weng, Wanjiang, Wang, Junbo, Zhao, Fang, Xie, Guo-Sen, Geng, Xin, Wang, Liang
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916905005088768
author Wang, Hongsong
Weng, Wanjiang
Wang, Junbo
Zhao, Fang
Xie, Guo-Sen
Geng, Xin
Wang, Liang
author_facet Wang, Hongsong
Weng, Wanjiang
Wang, Junbo
Zhao, Fang
Xie, Guo-Sen
Geng, Xin
Wang, Liang
contents Human action understanding serves as a foundational pillar in the field of intelligent motion perception. Skeletons serve as a modality- and device-agnostic representation for human modeling, and skeleton-based action understanding has potential applications in humanoid robot control and interaction. \RED{However, existing works often lack the scalability and generalization required to handle diverse action understanding tasks. There is no skeleton foundation model that can be adapted to a wide range of action understanding tasks}. This paper presents a Unified Skeleton-based Dense Representation Learning (USDRL) framework, which serves as a foundational model for skeleton-based human action understanding. USDRL consists of a Transformer-based Dense Spatio-Temporal Encoder (DSTE), Multi-Grained Feature Decorrelation (MG-FD), and Multi-Perspective Consistency Training (MPCT). The DSTE module adopts two parallel streams to learn temporal dynamic and spatial structure features. The MG-FD module collaboratively performs feature decorrelation across temporal, spatial, and instance domains to reduce dimensional redundancy and enhance information extraction. The MPCT module employs both multi-view and multi-modal self-supervised consistency training. The former enhances the learning of high-level semantics and mitigates the impact of low-level discrepancies, while the latter effectively facilitates the learning of informative multimodal features. We perform extensive experiments on 25 benchmarks across across 9 skeleton-based action understanding tasks, covering coarse prediction, dense prediction, and transferred prediction. Our approach significantly outperforms the current state-of-the-art methods. We hope that this work would broaden the scope of research in skeleton-based action understanding and encourage more attention to dense prediction tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2508_12586
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Foundation Model for Skeleton-Based Human Action Understanding
Wang, Hongsong
Weng, Wanjiang
Wang, Junbo
Zhao, Fang
Xie, Guo-Sen
Geng, Xin
Wang, Liang
Computer Vision and Pattern Recognition
Human action understanding serves as a foundational pillar in the field of intelligent motion perception. Skeletons serve as a modality- and device-agnostic representation for human modeling, and skeleton-based action understanding has potential applications in humanoid robot control and interaction. \RED{However, existing works often lack the scalability and generalization required to handle diverse action understanding tasks. There is no skeleton foundation model that can be adapted to a wide range of action understanding tasks}. This paper presents a Unified Skeleton-based Dense Representation Learning (USDRL) framework, which serves as a foundational model for skeleton-based human action understanding. USDRL consists of a Transformer-based Dense Spatio-Temporal Encoder (DSTE), Multi-Grained Feature Decorrelation (MG-FD), and Multi-Perspective Consistency Training (MPCT). The DSTE module adopts two parallel streams to learn temporal dynamic and spatial structure features. The MG-FD module collaboratively performs feature decorrelation across temporal, spatial, and instance domains to reduce dimensional redundancy and enhance information extraction. The MPCT module employs both multi-view and multi-modal self-supervised consistency training. The former enhances the learning of high-level semantics and mitigates the impact of low-level discrepancies, while the latter effectively facilitates the learning of informative multimodal features. We perform extensive experiments on 25 benchmarks across across 9 skeleton-based action understanding tasks, covering coarse prediction, dense prediction, and transferred prediction. Our approach significantly outperforms the current state-of-the-art methods. We hope that this work would broaden the scope of research in skeleton-based action understanding and encourage more attention to dense prediction tasks.
title Foundation Model for Skeleton-Based Human Action Understanding
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.12586