MEBM-Phoneme: Multi-scale Enhanced BrainMagic for End-to-End MEG Phoneme Classification

TL;DR

MEBM-Phoneme is a novel multi-scale neural decoder that enhances MEG phoneme classification by integrating hierarchical temporal features and robust training strategies, achieving high accuracy and generalization.

Contribution

The paper introduces a multi-scale convolutional architecture with attention and training techniques tailored for MEG phoneme decoding, improving over existing methods.

Findings

Achieves competitive accuracy on LibriBrain dataset

Demonstrates robustness to session-specific shifts

Validates effectiveness of hierarchical temporal modeling

Abstract

We propose MEBM-Phoneme, a multi-scale enhanced neural decoder for phoneme classification from non-invasive magnetoencephalography (MEG) signals. Built upon the BrainMagic backbone, MEBM-Phoneme integrates a short-term multi-scale convolutional module to augment the native mid-term encoder, with fused representations via depthwise separable convolution for efficient cross-scale integration. A convolutional attention layer dynamically weights temporal dependencies to refine feature aggregation. To address class imbalance and session-specific distributional shifts, we introduce a stacking-based local validation set alongside weighted cross-entropy loss and random temporal augmentation. Comprehensive evaluations on LibriBrain Competition 2025 Track2 demonstrate robust generalization, achieving competitive phoneme decoding accuracy on the validation and official test leaderboard. These results underscore the value of hierarchical temporal modeling and training stabilization for advancing MEG-based speech perception analysis.