# Measuring Robustness of Speech Recognition from MEG Signals Under Distribution Shift

arXiv: https://arxiv.org/abs/2604.04129 · HTML: https://mlsanity.com/paper/2604.04129/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 15 of its 31 references that are in the map.
- First posted on arXiv 5 Apr 2026.

## Builds on (most cited first, 15 of 15)

- [SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition](https://mlsanity.com/paper/1904.08779/) (2019)
- [Instance Normalization: The Missing Ingredient for Fast Stylization](https://mlsanity.com/paper/1607.08022/) (2016)
- [Deep learning for time series classification: a review](https://mlsanity.com/paper/1809.04356/) (2018)
- [LSTM Fully Convolutional Networks for Time Series Classification](https://mlsanity.com/paper/1709.05206/) (2017)
- [Decoding speech perception from non-invasive brain recordings](https://mlsanity.com/paper/2208.12266/) (2022)
- [Deep Learning for Time-Series Analysis](https://mlsanity.com/paper/1701.01887/) (2017)
- [Are EEG-to-Text Models Working?](https://mlsanity.com/paper/2405.06459/) (2024)
- [Deep Gated Recurrent and Convolutional Network Hybrid Model for Univariate Time Series Classification](https://mlsanity.com/paper/1812.07683/) (2018)
- [Domain Adaptation via Teacher-Student Learning for End-to-End Speech Recognition](https://mlsanity.com/paper/2001.01798/) (2020)
- [The Brain's Bitter Lesson: Scaling Speech Decoding With Self-Supervised Learning](https://mlsanity.com/paper/2406.04328/) (2024)
- [LibriBrain: Over 50 Hours of Within-Subject MEG to Improve Speech Decoding Methods at Scale](https://mlsanity.com/paper/2506.02098/) (2025)
- [The 2025 PNPL Competition: Speech Detection and Phoneme Classification in the LibriBrain Dataset](https://mlsanity.com/paper/2506.10165/) (2025)
- [Resolving Domain Shift For Representations Of Speech In Non-Invasive Brain Recordings](https://mlsanity.com/paper/2410.19986/) (2024)
- [MEGConformer: Conformer-Based MEG Decoder for Robust Speech and Phoneme Classification](https://mlsanity.com/paper/2512.01443/) (2025)
- [Multi-Subdomain Adversarial Network for Cross-Subject EEG-based Emotion Recognition](https://mlsanity.com/paper/2308.14059/) (2023)

## Abstract (arXiv)

This study investigates robust speech-related decoding from non-invasive MEG signals using the LibriBrain phoneme-classification benchmark from the 2025 PNPL competition. We compare residual convolutional neural networks (CNNs), an STFT-based CNN, and a CNN--Transformer hybrid, while also examining the effects of group averaging, label balancing, repeated grouping, normalization strategies, and data augmentation. Across our in-house implementations, preprocessing and data-configuration choices matter more than additional architectural complexity, among which instance normalization emerges as the most influential modification for generalization. The strongest of our own models, a CNN with group averaging, label balancing, repeated grouping, and instance normalization, achieves 60.95% F1-macro on the test split, compared with 39.53% for the plain CNN baseline. However, most of our models, without instance normalization, show substantial validation-to-test degradation, indicating that distribution shift induced by different normalization statistics is a major obstacle to generalization in our experiments. By contrast, MEGConformer maintains 64.09% F1-macro on both validation and test, and saliency-map analysis is qualitatively consistent with this contrast: weaker models exhibit more concentrated or repetitive phoneme-sensitive patterns across splits, whereas MEGConformer appears more distributed. Overall, the results suggest that improving the reliability of non-invasive phoneme decoding will likely require better handling of normalization-related distribution shift while also addressing the challenge of single-trial decoding.