New Synthetic Goldmine: Hand Joint Angle-Driven EMG Data Generation Framework for Micro-Gesture Recognition

2025-09-27

AI Summary

To address three key challenges in electromyography (EMG)-based gesture recognition—scarcity of labeled data, poor cross-subject generalization, and limited adaptability to unseen gestures—this paper proposes SeqEMG-GAN, a conditional generative framework that synthesizes high-fidelity, physiologically plausible EMG signals from hand joint-angle sequences. Its core innovations include the Ang2Gist semantic mapping module and a two-level context-aware architecture, enabling interpretable and semantically consistent translation from gesture semantics to EMG patterns. The model employs an angle-encoder-driven sequential GAN structure, jointly optimizing a deep convolutional generator and discriminator. A classifier trained solely on synthetic data achieves 55.71% accuracy; when real and synthetic data are fused, accuracy rises to 60.53%, surpassing the real-data-only baseline by 2.76%. These results demonstrate SeqEMG-GAN’s effectiveness in data augmentation and cross-user generalization.

Technology Category

• Natural Language Processing: Generation
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Electromyography (EMG)-based gesture recognition has emerged as a promising approach for human-computer interaction. However, its performance is often limited by the scarcity of labeled EMG data, significant cross-user variability, and poor generalization to unseen gestures. To address these challenges, we propose SeqEMG-GAN, a conditional, sequence-driven generative framework that synthesizes high-fidelity EMG signals from hand joint angle sequences. Our method introduces a context-aware architecture composed of an angle encoder, a dual-layer context encoder featuring the novel Ang2Gist unit, a deep convolutional EMG generator, and a discriminator, all jointly optimized via adversarial learning. By conditioning on joint kinematic trajectories, SeqEMG-GAN is capable of generating semantically consistent EMG sequences, even for previously unseen gestures, thereby enhancing data diversity and physiological plausibility. Experimental results show that classifiers trained solely on synthetic data experience only a slight accuracy drop (from 57.77% to 55.71%). In contrast, training with a combination of real and synthetic data significantly improves accuracy to 60.53%, outperforming real-only training by 2.76%. These findings demonstrate the effectiveness of our framework,also achieves the state-of-art performance in augmenting EMG datasets and enhancing gesture recognition performance for applications such as neural robotic hand control, AI/AR glasses, and gesture-based virtual gaming systems.

Problem

• Generating synthetic EMG data from hand joint angles
• Addressing scarcity and variability in EMG gesture recognition
• Enhancing generalization for unseen gestures with SeqEMG-GAN

Innovation

• Generates EMG signals from hand joint angles
• Uses conditional GAN with context-aware architecture
• Enhances gesture recognition with synthetic data augmentation