ExoGait-MS: Learning Periodic Dynamics with Multi-Scale Graph Network for Exoskeleton Gait Recognition

TL;DR

This paper tackles the challenge of personalized gait recognition for exoskeleton control, where standardized gaits fail to accommodate individual locomotor differences. It introduces a dual-branch framework that combines a nonlinear periodic dynamics learning module with a cross-scale multi-scale dense graph convolutional network, fused through a spatiotemporal layer to capture both temporal and spatial gait features. The authors validate their approach on a self-collected dataset (24 participants, 5 environments) and report a peak accuracy of 94.34%, outperforming state-of-the-art methods by 3.77%. The work has practical impact by enabling real-time, user-specific exoskeleton control and safer, more effective rehabilitation through precise gait understanding and adaptive assistance.

Abstract

Current exoskeleton control methods often face challenges in delivering personalized treatment. Standardized walking gaits can lead to patient discomfort or even injury. Therefore, personalized gait is essential for the effectiveness of exoskeleton robots, as it directly impacts their adaptability, comfort, and rehabilitation outcomes for individual users. To enable personalized treatment in exoskeleton-assisted therapy and related applications, accurate recognition of personal gait is crucial for implementing tailored gait control. The key challenge in gait recognition lies in effectively capturing individual differences in subtle gait features caused by joint synergy, such as step frequency and step length. To tackle this issue, we propose a novel approach, which uses Multi-Scale Global Dense Graph Convolutional Networks (GCN) in the spatial domain to identify latent joint synergy patterns. Moreover, we propose a Gait Non-linear Periodic Dynamics Learning module to effectively capture the periodic characteristics of gait in the temporal domain. To support our individual gait recognition task, we have constructed a comprehensive gait dataset that ensures both completeness and reliability. Our experimental results demonstrate that our method achieves an impressive accuracy of 94.34% on this dataset, surpassing the current state-of-the-art (SOTA) by 3.77%. This advancement underscores the potential of our approach to enhance personalized gait control in exoskeleton-assisted therapy.

Figures (4)

• Figure 1: The architecture of our proposed gait recognition framework consisting of two parallel branches. The top branch learns temporal gait dynamics via an encoder-decoder that captures nonlinear periodic patterns in a low-dimensional latent space, while the bottom branch extracts spatial gait features through cross-scale global graph learning, explicitly modeling joint-wise dependencies.
• Figure 2: Pipeline for extracting 3D gait information from an input video using a 2D pose estimator and a 2D-to-3D lifting model.
• Figure 3: This is an example of detection results with 3D keypoints for different individuals.
• Figure 4: Visualization of the learned adjacency matrix across different graph layers.