Beyond Binary Contrast: Modeling Continuous Skeleton Action Spaces with Transitional Anchors

2026-04-20

AI Summary

This work addresses the limitation of existing binary contrastive self-supervised methods, which neglect the intrinsic continuity of human actions, leading to fragmented representations and rigid category boundaries. To overcome this, we propose TranCLR, a novel framework that explicitly models action continuity through action transition anchors and introduces a Multi-level Geometric Manifold Calibration (MGMC) mechanism to adaptively refine the action manifold structure. By integrating Action Transition Anchor Construction (ATAC) with self-supervised contrastive learning, TranCLR learns smooth, discriminative skeleton-based action representations endowed with uncertainty awareness. Extensive experiments on NTU RGB+D, NTU RGB+D 120, and PKU-MMD benchmarks demonstrate that TranCLR achieves state-of-the-art performance in both classification accuracy and calibration quality.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Motion & Tracking
• Reasoning under Uncertainty: Uncertainty Representations

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Self-supervised contrastive learning has emerged as a powerful paradigm for skeleton-based action recognition by enforcing consistency in the embedding space. However, existing methods rely on binary contrastive objectives that overlook the intrinsic continuity of human motion, resulting in fragmented feature clusters and rigid class boundaries. To address these limitations, we propose TranCLR, a Transitional anchor-based Contrastive Learning framework that captures the continuous geometry of the action space. Specifically, the proposed Action Transitional Anchor Construction (ATAC) explicitly models the geometric structure of transitional states to enhance the model's perception of motion continuity. Building upon these anchors, a Multi-Level Geometric Manifold Calibration (MGMC) mechanism is introduced to adaptively calibrate the action manifold across multiple levels of continuity, yielding a smoother and more discriminative representation space. Extensive experiments on the NTU RGB+D, NTU RGB+D 120 and PKU-MMD datasets demonstrate that TranCLR achieves superior accuracy and calibration performance, effectively learning continuous and uncertainty-aware skeleton representations. The code is available at https://github.com/Philchieh/TranCLR.

Problem

• skeleton-based action recognition
• contrastive learning
• motion continuity
• feature clustering
• class boundaries

Innovation

• Transitional Anchors
• Contrastive Learning
• Action Continuity
• Geometric Manifold Calibration
• Skeleton-based Action Recognition