A Robust Multi-Scale Framework with Test-Time Adaptation for sEEG-Based Speech Decoding

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wang, Suli, Li, Yang-yang, Cai, Siqi, Li, Haizhou
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918150551896064
author Wang, Suli
Li, Yang-yang
Cai, Siqi
Li, Haizhou
author_facet Wang, Suli
Li, Yang-yang
Cai, Siqi
Li, Haizhou
contents Decoding speech from stereo-electroencephalography (sEEG) signals has emerged as a promising direction for brain-computer interfaces (BCIs). Its clinical applicability, however, is limited by the inherent non-stationarity of neural signals, which causes domain shifts between training and testing, undermining decoding reliability. To address this challenge, a two-stage framework is proposed for enhanced robustness. First, a multi-scale decomposable mixing (MDM) module is introduced to model the hierarchical temporal dynamics of speech production, learning stable multi-timescale representations from sEEG signals. Second, a source-free online test-time adaptation (TTA) method performs entropy minimization to adapt the model to distribution shifts during inference. Evaluations on the public DU-IN spoken word decoding benchmark show that the approach outperforms state-of-the-art models, particularly in challenging cases. This study demonstrates that combining invariant feature learning with online adaptation is a principled strategy for developing reliable BCI systems. Our code is available at https://github.com/lyyi599/MDM-TENT.
format Preprint
id arxiv_https___arxiv_org_abs_2509_24700
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Robust Multi-Scale Framework with Test-Time Adaptation for sEEG-Based Speech Decoding
Wang, Suli
Li, Yang-yang
Cai, Siqi
Li, Haizhou
Human-Computer Interaction
Decoding speech from stereo-electroencephalography (sEEG) signals has emerged as a promising direction for brain-computer interfaces (BCIs). Its clinical applicability, however, is limited by the inherent non-stationarity of neural signals, which causes domain shifts between training and testing, undermining decoding reliability. To address this challenge, a two-stage framework is proposed for enhanced robustness. First, a multi-scale decomposable mixing (MDM) module is introduced to model the hierarchical temporal dynamics of speech production, learning stable multi-timescale representations from sEEG signals. Second, a source-free online test-time adaptation (TTA) method performs entropy minimization to adapt the model to distribution shifts during inference. Evaluations on the public DU-IN spoken word decoding benchmark show that the approach outperforms state-of-the-art models, particularly in challenging cases. This study demonstrates that combining invariant feature learning with online adaptation is a principled strategy for developing reliable BCI systems. Our code is available at https://github.com/lyyi599/MDM-TENT.
title A Robust Multi-Scale Framework with Test-Time Adaptation for sEEG-Based Speech Decoding
topic Human-Computer Interaction
url https://arxiv.org/abs/2509.24700