脑电基础模型能否迁移至语音？显性与想象语音解码的基准测试
Do EEG Foundation Models Transfer to Speech? A Benchmark on Overt and Imagined Speech Decoding
浏览论文内容
中文总结 AI 辅助
本文开展首个脑电基础模型语音解码基准测试，对比LaBraM等模型与EEGNet等基线，发现通用脑电预训练未在语音任务上展现一致优势，为语音专用基础模型研究提供依据。
中文摘要 AI 辅助
基于数千小时数据预训练的脑电基础模型，在运动想象、癫痫检测、睡眠分期和情绪识别等任务上，较任务特定网络展现出显著性能提升，但它们向语音解码的迁移——这无疑是要求最高的非侵入式BCI应用——仍未得到验证。本文开展了首个针对语音解码的脑电基础模型系统性基准测试，对比强大的卷积基线模型，使用两个语料库：UGR-MINDVOICE（显性与隐性伊比利亚西班牙语）和BCI Competition 2020 Track 3（想象语音）。在统一的预处理与微调协议下，我们对比了两个基础模型（LaBraM、EEGMamba）与三个已确立的基线模型（EEGNet、ShallowFBCSPNet、EEGConformer）。大规模脑电预训练在语音任务上未展现出对1.6万参数卷积神经网络的一致优势，表明当前通用脑电预训练尚未能迁移至语音生成，为语音专用基础模型的研究提供了动机。
英文摘要
EEG foundation models pretrained on thousands of hours have shown large gains over task-specific networks for motor imagery, seizure detection, sleep staging, and emotion recognition, but their transfer to speech decoding - arguably the most demanding non-invasive BCI application - remains untested. We present the first systematic benchmark of EEG foundation models against strong convolutional baselines for speech decoding, using two corpora: UGR-MINDVOICE (overt and covert Iberian Spanish) and BCI Competition 2020 Track 3 (imagined speech). We compare two foundation models (LaBraM, EEGMamba) against three established baselines (EEGNet, ShallowFBCSPNet, EEGConformer) under a unified preprocessing and fine-tuning protocol. Large-scale EEG pretraining yields no consistent advantage over a 16K-parameter CNN on speech tasks, indicating that current general-purpose EEG pretraining does not yet transfer to speech production and motivating speech-specific foundation models.
发表机构
- University of Granada, Spain（格拉纳达大学）
- Research Centre for Information and Communication Technologies (CITIC-UGR), Spain（信息与通信技术研究中心）
- University of Chouaib Doukkali, Morocco（舒艾卜·杜卡利大学）
- Brain, Mind, and Behavior Research Center (CIMCYC), University of Granada, Spain（大脑、心智与行为研究中心）
机构由 AI 辅助整理，请以论文原文为准。
