Does EEG Foundation Models Transfer to Speech? A Benchmark on Overt and Imagined Speech Decoding

2026/07/29 by Owais Mujtaba Khanday, Mohamed Baha Ben Ticha, Sanae Belfrouh +2
Computer Science · #cs.SD

paper · pdf

6 pages, 1 figure, 3 tables, submitted to IberSPEECH 2026

arxiv created 2026/07/29 · arxiv updated 2026/07/31

Abstract

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.

Citations

• Neural Decoding of Overt Speech from ECoG Using Vision Transformers and Contrastive Representation Learning
• BrainECHO: Semantic Brain Signal Decoding through Vector-Quantized Spectrogram Reconstruction for Whisper-Enhanced Text Generation
• EEG Conformer: Convolutional Transformer for EEG Decoding and Visualization
• Decoding speech perception from non-invasive brain recordings
• Deep learning with convolutional neural networks for EEG decoding and visualization
• Adam: A Method for Stochastic Optimization
• BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Related

• A Penny for Your Thoughts: Decoding Speech from Inexpensive Brain Signals
• EEG Foundation Models for BCI Learn Diverse Features of Electrophysiology
• BENDR: using transformers and a contrastive self-supervised learning\n task to learn from massive amounts of EEG data
• Scaling Law in Neural Data: Non-Invasive Speech Decoding with 175 Hours of EEG Data
• Towards Unified Neural Decoding of Perceived, Spoken and Imagined Speech from EEG Signals