BrainWhisperer: Leveraging Large-Scale ASR Models for Neural Speech Decoding

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Boccato, Tommaso, Olak, Michal, Ferrante, Matteo
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910052479139840
author Boccato, Tommaso
Olak, Michal
Ferrante, Matteo
author_facet Boccato, Tommaso
Olak, Michal
Ferrante, Matteo
contents Decoding continuous speech from intracortical recordings is a central challenge for brain-computer interfaces (BCIs), with transformative potential for individuals with conditions that impair their ability to speak. While recent microelectrode array (MEA) decoders achieve impressive accuracy, their performance is fundamentally limited by the small size of existing datasets, they remain brittle to session-to-session variability, and their ability to generalize across participants remains unexplored. We introduce BrainWhisperer, a neural speech decoder that integrates high-resolution MEA recordings with a large pretrained automatic speech recognition (ASR) model. Building on interpretability findings showing that Whisper's encoder learns phoneme-selective representations with localized attention, we train a customized version of Whisper, modified to process neural features, using a hybrid objective that combines CTC loss on phonemes--predicted from the third encoder layer--and cross-entropy loss on word tokens. We introduce domain-informed modifications including windowed self-attention to capture articulatory continuity, hierarchical month/day-specific low-rank projections to address non-stationarity, and subject-specific embedders enabling cross-subject training. Evaluated on a publicly available MEA dataset (Card et al.), BrainWhisperer matches or outperforms prior state-of-the-art decoders. Critically, cross-dataset training improves performance even on individual datasets without fine-tuning, demonstrating unprecedented generalization. The model supports dual decoding paths: a high-accuracy phoneme-based path with external language model rescoring, and a fast direct text generation path enabling sub-100ms inference with minimal hardware requirements.
format Preprint
id arxiv_https___arxiv_org_abs_2603_13321
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle BrainWhisperer: Leveraging Large-Scale ASR Models for Neural Speech Decoding
Boccato, Tommaso
Olak, Michal
Ferrante, Matteo
Audio and Speech Processing
Sound
Decoding continuous speech from intracortical recordings is a central challenge for brain-computer interfaces (BCIs), with transformative potential for individuals with conditions that impair their ability to speak. While recent microelectrode array (MEA) decoders achieve impressive accuracy, their performance is fundamentally limited by the small size of existing datasets, they remain brittle to session-to-session variability, and their ability to generalize across participants remains unexplored. We introduce BrainWhisperer, a neural speech decoder that integrates high-resolution MEA recordings with a large pretrained automatic speech recognition (ASR) model. Building on interpretability findings showing that Whisper's encoder learns phoneme-selective representations with localized attention, we train a customized version of Whisper, modified to process neural features, using a hybrid objective that combines CTC loss on phonemes--predicted from the third encoder layer--and cross-entropy loss on word tokens. We introduce domain-informed modifications including windowed self-attention to capture articulatory continuity, hierarchical month/day-specific low-rank projections to address non-stationarity, and subject-specific embedders enabling cross-subject training. Evaluated on a publicly available MEA dataset (Card et al.), BrainWhisperer matches or outperforms prior state-of-the-art decoders. Critically, cross-dataset training improves performance even on individual datasets without fine-tuning, demonstrating unprecedented generalization. The model supports dual decoding paths: a high-accuracy phoneme-based path with external language model rescoring, and a fast direct text generation path enabling sub-100ms inference with minimal hardware requirements.
title BrainWhisperer: Leveraging Large-Scale ASR Models for Neural Speech Decoding
topic Audio and Speech Processing
Sound
url https://arxiv.org/abs/2603.13321