Spatio-Temporal Joint Density Driven Learning for Skeleton-Based Action Recognition

2025-05-29

IEEE Transactions on Biometrics Behavior and Identity Science

AI Summary

Existing skeleton-based action recognition methods often neglect the discriminative synergy between dynamic (moving) and static (supporting) joints, limiting representation learning. Method: This paper introduces Spatio-Temporal Joint Density (STJD), a novel metric that explicitly models the co-evolutionary relationship between dynamic and static joints and adaptively identifies discriminative “principal joints.” Building upon STJD, we propose STJD-CL, a contrastive learning strategy, and STJD-MP, a reconstruction-augmented framework, enabling motion-static interaction-driven self-supervised representation learning. Our approach employs a graph convolutional network backbone. Results: On NTU RGB+D 120, it achieves state-of-the-art performance, outperforming prior methods by 3.5% and 3.6% in the cross-subject and cross-setup protocols, respectively. It also consistently surpasses all competitors on NTU-60 and PKUMMD. These results demonstrate the effectiveness and generalizability of modeling dynamic-static joint synergy for skeleton-based action recognition.

Technology Category

• Computer Vision: Motion & Tracking
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Machine Learning: Representation Learning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Traditional approaches in unsupervised or self supervised learning for skeleton-based action classification have concentrated predominantly on the dynamic aspects of skeletal sequences. Yet, the intricate interaction between the moving and static elements of the skeleton presents a rarely tapped discriminative potential for action classification. This paper introduces a novel measurement, referred to as spatial-temporal joint density (STJD), to quantify such interaction. Tracking the evolution of this density throughout an action can effectively identify a subset of discriminative moving and/or static joints termed"prime joints"to steer self-supervised learning. A new contrastive learning strategy named STJD-CL is proposed to align the representation of a skeleton sequence with that of its prime joints while simultaneously contrasting the representations of prime and nonprime joints. In addition, a method called STJD-MP is developed by integrating it with a reconstruction-based framework for more effective learning. Experimental evaluations on the NTU RGB+D 60, NTU RGB+D 120, and PKUMMD datasets in various downstream tasks demonstrate that the proposed STJD-CL and STJD-MP improved performance, particularly by 3.5 and 3.6 percentage points over the state-of-the-art contrastive methods on the NTU RGB+D 120 dataset using X-sub and X-set evaluations, respectively.

Problem

• Quantify spatial-temporal joint density for skeleton action recognition
• Identify discriminative prime joints for self-supervised learning
• Improve contrastive learning performance on skeleton-based datasets

Innovation

• Spatial-temporal joint density (STJD) measures skeleton interaction
• STJD-CL aligns skeleton with prime joints contrastively
• STJD-MP integrates reconstruction for enhanced learning