WaveFormer: A Lightweight Transformer Model for sEMG-based Gesture Recognition

TL;DR

WaveFormer is a lightweight transformer model that effectively recognizes sEMG gestures by combining time and frequency features, achieving high accuracy and real-time performance on resource-limited devices.

Contribution

The paper introduces WaveFormer, a novel compact transformer architecture with a learnable wavelet transform for efficient sEMG gesture recognition.

Findings

Achieves 95% accuracy on EPN612 dataset

Contains only 3.1 million parameters

Enables real-time inference with 6.75 ms latency

Abstract

Human-machine interaction, particularly in prosthetic and robotic control, has seen progress with gesture recognition via surface electromyographic (sEMG) signals.However, classifying similar gestures that produce nearly identical muscle signals remains a challenge, often reducing classification accuracy. Traditional deep learning models for sEMG gesture recognition are large and computationally expensive, limiting their deployment on resource-constrained embedded systems. In this work, we propose WaveFormer, a lightweight transformer-based architecture tailored for sEMG gesture recognition. Our model integrates time-domain and frequency-domain features through a novel learnable wavelet transform, enhancing feature extraction. In particular, the WaveletConv module, a multi-level wavelet decomposition layer with depthwise separable convolution, ensures both efficiency and compactness. With just 3.1 million parameters, WaveFormer achieves 95% classification accuracy on the EPN612 dataset, outperforming larger models. Furthermore, when profiled on a laptop equipped with an Intel CPU, INT8 quantization achieves real-time deployment with a 6.75 ms inference latency.