Articulatory Feature Prediction from Surface EMG during Speech Production

TL;DR

This paper introduces a novel model combining convolutional and Transformer layers to predict articulatory features from surface EMG signals, enabling the decoding of intelligible speech waveforms and optimizing electrode placement.

Contribution

The study presents the first method to decode speech waveforms directly from surface EMG via predicted articulatory features, integrating advanced neural network components.

Findings

Achieves approximately 0.9 correlation in articulatory feature prediction

Successfully decodes intelligible speech waveforms from EMG signals

Provides insights into electrode placement for improved predictability

Abstract

We present a model for predicting articulatory features from surface electromyography (EMG) signals during speech production. The proposed model integrates convolutional layers and a Transformer block, followed by separate predictors for articulatory features. Our approach achieves a high prediction correlation of approximately 0.9 for most articulatory features. Furthermore, we demonstrate that these predicted articulatory features can be decoded into intelligible speech waveforms. To our knowledge, this is the first method to decode speech waveforms from surface EMG via articulatory features, offering a novel approach to EMG-based speech synthesis. Additionally, we analyze the relationship between EMG electrode placement and articulatory feature predictability, providing knowledge-driven insights for optimizing EMG electrode configurations. The source code and decoded speech samples are publicly available.