Context Informed Incremental Learning Improves Myoelectric Control Performance in Virtual Reality Object Manipulation Tasks

2025-05-09

AI Summary

EMG-based gesture recognition systems exhibit strong performance in controlled laboratory settings but suffer significant degradation in real-world, goal-directed tasks—such as VR object manipulation—due to dynamic behavioral context shifts, undermining long-term usability. To address this, we propose Context-guided Incremental Learning (CIIL), the first framework to integrate task-semantic context into EMG-driven VR interaction. Our method enables online, context-aware adaptation of the EMG classifier by dynamically fusing real-time user behavioral context, thereby departing from conventional static offline training paradigms. Experimental evaluation demonstrates substantial improvements in task success rate and operational efficiency, alongside a 7.1% reduction in subjective cognitive load. This work establishes the critical role of context-aware online learning in enhancing the robustness and user experience of deployed EMG systems, offering a novel paradigm for adaptive wearable human–computer interaction.

Technology Category

• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Machine Learning: Online Learning & Bandits
• Intelligent Robots: Behavior Learning & Control

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience

Abstract

Electromyography (EMG)-based gesture recognition is a promising approach for designing intuitive human-computer interfaces. However, while these systems typically perform well in controlled laboratory settings, their usability in real-world applications is compromised by declining performance during real-time control. This decline is largely due to goal-directed behaviors that are not captured in static, offline scenarios. To address this issue, we use extit{Context Informed Incremental Learning} (CIIL) - marking its first deployment in an object-manipulation scenario - to continuously adapt the classifier using contextual cues. Nine participants without upper limb differences completed a functional task in a virtual reality (VR) environment involving transporting objects with life-like grips. We compared two scenarios: one where the classifier was adapted in real-time using contextual information, and the other using a traditional open-loop approach without adaptation. The CIIL-based approach not only enhanced task success rates and efficiency, but also reduced the perceived workload by 7.1 %, despite causing a 5.8 % reduction in offline classification accuracy. This study highlights the potential of real-time contextualized adaptation to enhance user experience and usability of EMG-based systems for practical, goal-oriented applications, crucial elements towards their long-term adoption. The source code for this study is available at: https://github.com/BiomedicalITS/ciil-emg-vr.

Problem

• Improving EMG-based gesture recognition in real-time control
• Addressing performance decline in real-world EMG applications
• Enhancing usability of EMG systems with contextual adaptation

Innovation

• Context Informed Incremental Learning adapts classifier
• Uses contextual cues for real-time control
• Enhances task success and reduces workload