MEG-to-MEG Transfer Learning and Cross-Task Speech/Silence Detection with Limited Data

TL;DR

This study demonstrates that transfer learning significantly enhances MEG-based speech decoding across perception and production tasks with limited data, revealing shared neural representations.

Contribution

First application of transfer learning and cross-task decoding in MEG speech models, enabling effective decoding with minimal data across different speech tasks.

Findings

Transfer learning improves decoding accuracy by 1-6%.

Models trained on production data decode listening above chance.

Shared neural representations are confirmed across tasks.

Abstract

Data-efficient neural decoding is a central challenge for speech brain-computer interfaces. We present the first demonstration of transfer learning and cross-task decoding for MEG-based speech models spanning perception and production. We pre-train a Conformer-based model on 50 hours of single-subject listening data and fine-tune on just 5 minutes per subject across 18 participants. Transfer learning yields consistent improvements, with in-task accuracy gains of 1-4% and larger cross-task gains of up to 5-6%. Not only does pre-training improve performance within each task, but it also enables reliable cross-task decoding between perception and production. Critically, models trained on speech production decode passive listening above chance, confirming that learned representations reflect shared neural processes rather than task-specific motor activity.