The Brain's Bitter Lesson: Scaling Speech Decoding With Self-Supervised Learning
This work addresses the problem of leveraging heterogeneous brain recordings for speech decoding, enabling scaling beyond current limitations for researchers and clinicians.
The paper tackled the challenge of generalizing speech decoding models across subjects and datasets by developing a self-supervised learning approach that scales to nearly 400 hours of MEG data and 900 subjects, achieving improvements of 15-27% over state-of-the-art models and matching surgical decoding performance with non-invasive data.
The past few years have seen remarkable progress in the decoding of speech from brain activity, primarily driven by large single-subject datasets. However, due to individual variation, such as anatomy, and differences in task design and scanning hardware, leveraging data across subjects and datasets remains challenging. In turn, the field has not benefited from the growing number of open neural data repositories to exploit large-scale deep learning. To address this, we develop neuroscience-informed self-supervised objectives, together with an architecture, for learning from heterogeneous brain recordings. Scaling to nearly 400 hours of MEG data and 900 subjects, our approach shows generalisation across participants, datasets, tasks, and even to novel subjects. It achieves improvements of 15-27% over state-of-the-art models and matches surgical decoding performance with non-invasive data. These advances unlock the potential for scaling speech decoding models beyond the current frontier.