arXiv Machine Learning

GRAFT-EEG: Gauge-Aware Physical Field Transport for Cross-Dataset Motor-Imagery EEG

The paper introduces GRAFT-EEG, a method that removes the active‑set common‑reference gauge and transports each native EEG mask through a template motor leadfield, enabling target‑statistics‑blind transfer for cross‑dataset motor‑imagery decoding. Using a source‑fitted spectral classifier and a bounded state‑space‑model residual, the approach achieves higher balanced accuracy (60.865%) than a strict‑clean baseline (52.684%) across five leave‑one‑dataset‑out targets, and outperforms several control conditions in post‑hoc tests. The study also reports positive results in a separate 60‑participant same‑paradigm evaluation, though the dynamic‑static comparison was inconclusive. "whyItMatters":"The method demonstrates that accounting for observation‑operator changes can substantially improve cross‑dataset EEG decoding performance, offering a more robust approach for motor‑imagery applications."

arXiv Machine Learning

Sep 2

Lightweight Adaptation of EEG Foundation Models for Stroke Motor Imagery Decoding: Domain Shift and Subject-Level Robustness

The study investigates whether Low‑Rank Adaptation (LoRA) can adapt three pretrained EEG foundation models—LaBraM‑base, REVE‑base, and REVE‑large—for binary left‑ vs. right‑hand motor imagery decoding in stroke patients. Using subject‑wise five‑fold cross‑validation on the PhysioNet EEG Motor Movement/Imagery Dataset and a binary subset of the UET175 stroke dataset, LoRA significantly improved accuracy for LaBraM‑base (0.822) and REVE‑base (0.957) on the healthy cohort, but only REVE‑base LoRA achieved high performance (0.847±0.194) on stroke data, with a best mean accuracy of 0.952 in leave‑one‑subject‑out evaluation. The results demonstrate that healthy‑benchmark performance does not guarantee transfer to stroke EEG, highlighting the need for target‑domain adaptation and subject‑level assessment in rehabilitation BCIs.

By Anh T. Nguyen, Zihua Sun, Michelle J. Johnson

fine-tuning benchmarks

arXiv Machine Learning

Jul 28

Generative Augmentation for EEG Motor Imagery Classification: A Class-Conditional VAE with Cycle-Consistent Decoder Refinement

arXiv:2607. 22733v1 Announce Type: cross Abstract: We investigate whether a generative model can supply useful synthetic motor-imagery (MI) electroencephalography (EEG) trials that improve the accuracy of independent downstream classifiers.

By Matei Moldoveanu, Alain Sirois, Claire Ben Ali, Fabien Lotte, Florian Yger

arXiv Machine Learning

Aug 4

What EEG Foundation Models Encode: Dataset Identity and a Negative-Control Suite for Clinical Benchmarks

arXiv:2607. 24519v2 Announce Type: replace Abstract: Pretrained EEG foundation models are proposed for clinical decoding, but whether reported gains transfer across populations or survive negative controls is unclear.

By Marzieh Zare

rag fine-tuning benchmarks

arXiv AI

Oct 12

ThinkNet: Evolutionary Deep Ensembles for Efficient Motor-Imagery EEG Classification

ThinkNet is a validation‑controlled framework that uses train‑only normalization, evolutionary search guided by validation, and validation‑gated inference to build compact motor‑imagery EEG decoders that generalize to unseen subjects. In a fixed benchmark on the BCI Competition IV‑2a dataset, the best single model achieved 44.35 % accuracy with only 4.9 k parameters, while a compact ensemble of six families reached 43.31 % accuracy. Across a broader search, models with ≤25 k parameters outperformed larger ones after retraining, but validation reliability remains a bottleneck under subject shift.

By Abdul Basit, Saim Rehman, Muhammad Shafique

benchmarks

arXiv Machine Learning

Sep 22

Matched-Input Estimates Differ in Sign Across Architectures: Auditing EEG Foundation Models on Motor Imagery

arXiv:2609.23924v1 Announce Type: new Abstract: Pretrained EEG foundation models are increasingly proposed as general-purpose encoders for brain-computer interfaces, yet recent benchmarks disagree ab...

By Kevin Zhou, Sparsh Roy

fine-tuning benchmarks

arXiv Machine Learning

Jul 28

Bayesian Complete-Pooling in Cross-Subject Classification for Motor Imagery Electroencephalogram

arXiv:2607. 22980v1 Announce Type: new Abstract: Brain-computer interfaces (BCIs) have long sought calibration-free operation, but classifiers are typically benchmarked by discrimination alone, blind to whether predicted probabilities are well calibrated - a meaningful gap given nonstationary electroencephalogram (EEG) signals and the risk of overconfident point-estimate classifiers under distribution shift.

By Ethan Davis

arXiv Machine Learning

Jul 28

Subject-Level Heterogeneity in EEG Motor Imagery Decoding: A Large-Scale Benchmark and Portfolio-Based Reduction of the Search Space

arXiv:2607. 22778v1 Announce Type: cross Abstract: Robust EEG motor imagery decoding remains limited by strong inter-individual variability, making it difficult to identify pipelines that generalize across users.

By Paul Barbaste, Olivier Oullier, Xavier Vasques

benchmarks

arXiv AI

Jun 24

Average Rankings Mask Per-Subject Optimality: A Friedman-Nemenyi Benchmark of EEG Motor-Imagery BCI Decoders

arXiv:2606. 24394v1 Announce Type: cross Abstract: Electroencephalography (EEG) is the dominant non-invasive modality for brain-computer interfaces (BCIs), yet reliable decoding of motor imagery is hampered by inter- and intra-individual variability.

By Xavier Vasques, Paul Barbaste, Olivier Oullier

benchmarks

arXiv Machine Learning

Jul 7

Stacked LoRA for Subject-Adaptive EEG Foundation Models in Motor Imagery Decoding

arXiv:2607. 03094v1 Announce Type: new Abstract: Electroencephalography (EEG) decoding for brain-computer interfaces (BCIs) faces a major challenge: substantial inter-subject variability limits effective cross-subject generalization.

By Aymen Sarhane, Fouad Lbakali, Mouad Souissi, Jonathan Lys, Giulia Lioi

fine-tuning benchmarks

arXiv Machine Learning

Jun 8

The Identity Trap in EEG Foundation Models: A Diagnostic Audit

arXiv:2606. 06647v1 Announce Type: new Abstract: Objective.

By Jun-You Lin, Ying Choon Wu, Tzyy-Ping Jung

fine-tuning benchmarks

arXiv Computer Vision

Aug 27

Frequency-Decorrelated Temporal Ensembles for EEG--fNIRS Imagined-Handwriting Decoding

arXiv:2608.03176v2 Announce Type: replace Abstract: Imagined handwriting offers a temporally rich paradigm for non-invasive neural decoding, yet reliable recognition across unseen participants remain...

By Xiao Fan, Hongbin Guo, Yubo Han, Yi Zhang

multimodal

arXiv Machine Learning

Sep 16

MyoFlow: Anchor-Tied Rectified Flow for HD-sEMG Gesture Recognition Across Sessions and Subjects

MyoFlow introduces a discriminative flow-matching framework for high‑density surface electromyography (HD‑sEMG) gesture recognition that addresses distribution shifts caused by electrode re‑donning and physiological variability. By using a domain‑conditioned rectified flow to transport encoded windows toward gesture anchors, the method enables zero‑shot prediction without a separate classifier head. On the Hyser dataset, MyoFlow outperforms the strongest diffusion‑based baseline by 4.24 % in cross‑session accuracy and 6.37 % in cross‑subject accuracy, and achieves 91.71 % mean zero‑shot accuracy and 97.39 % mean few‑shot accuracy on the CEMHSEY dataset.

By Chenhao Wu, Dingjie Peng, Satoshi Funabashi, Satoshi Konishi, Wuqiang Yang, Hiroshi Onoda, Hironori Washizaki, Jiang Liu

diffusion robotics

Related stories