DynSTG-Mamba: Dynamic Spatio-Temporal Graph Mamba with Cross-Graph Knowledge Distillation for Gait Disorders Recognition

2025-03-17

AI Summary

To address the high memory overhead and limited capacity of existing Spatio-Temporal Graph Convolutional Networks (ST-GCNs) in modeling complex spatio-temporal dependencies for gait disorder recognition, this paper proposes the Spatio-Temporal Graph Mamba (STG-Mamba) framework and a cross-graph relational knowledge distillation mechanism. Our contributions are threefold: (1) a novel Dynamic Filtering Spatio-Temporal Graph Neural Network (DF-STGNN) that adaptively captures joint-wise temporal dynamics; (2) an STG-Mamba architecture tailored to skeletal biomechanics, integrating graph-structured priors with state-space modeling; and (3) a cross-graph relational knowledge distillation strategy enabling structured transfer of joint topology and motion patterns from teacher to student models. Evaluated on three clinical datasets—KOA-NM, PD-WALK, and ATAXIA—our method achieves state-of-the-art performance in Accuracy, F1-score, and Recall, while significantly reducing computational cost, thus offering both lightweight efficiency and high accuracy.

Technology Category

• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data
• Machine Learning: Graph-based Machine Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Gait disorder recognition plays a crucial role in the early diagnosis and monitoring of movement disorders. Existing approaches, including spatio-temporal graph convolutional networks (ST-GCNs), often face high memory demands and struggle to capture complex spatio-temporal dependencies, limiting their efficiency in clinical applications. To address these challenges, we introduce DynSTG-Mamba (Dynamic Spatio-Temporal Graph Mamba), a novel framework that combines DF-STGNN and STG-Mamba to enhance motion sequence modeling. The DF-STGNN incorporates a dynamic spatio-temporal filter that adaptively adjusts spatial connections between skeletal joints and temporal interactions across different movement phases. This approach ensures better feature propagation through dynamic graph structures by considering the hierarchical nature and dynamics of skeletal gait data. Meanwhile, STG-Mamba, an extension of Mamba adapted for skeletal motion data, ensures a continuous propagation of states, facilitating the capture of long-term dependencies while reducing computational complexity. To reduce the number of model parameters and computational costs while maintaining consistency, we propose Cross-Graph Relational Knowledge Distillation, a novel knowledge transfer mechanism that aligns relational information between teacher (large architecture) and student models (small architecture) while using shared memory. This ensures that the interactions and movement patterns of the joints are accurately preserved in the motion sequences. We validate our DynSTG-Mamba on KOA-NM, PD-WALK, and ATAXIA datasets, where it outperforms state-of-the-art approaches by achieving in terms of Accuracy, F1-score, and Recall. Our results highlight the efficiency and robustness of our approach, offering a lightweight yet highly accurate solution for automated gait analysis and movement disorder assessment.

Problem

• Enhances gait disorder recognition using dynamic spatio-temporal graph modeling.
• Reduces computational complexity while capturing long-term dependencies in motion data.
• Proposes cross-graph knowledge distillation for efficient and accurate gait analysis.

Innovation

• Dynamic spatio-temporal filter enhances feature propagation.
• STG-Mamba captures long-term dependencies efficiently.
• Cross-Graph Knowledge Distillation reduces computational costs.