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

• URL: http://arxiv.org/abs/2607.03094v1
• Date: Fri, 03 Jul 2026 08:31:57 GMT
• Title: Stacked LoRA for Subject-Adaptive EEG Foundation Models in Motor Imagery Decoding
• Abstract summary: We propose a framework that structurally decouples subject-invariant knowledge from subject-specific neural signatures.<n> Experiments on BCI Competition IV-2a, PhysioNet Motor Imagery, and the clinical Zuo2025 benchmark show that Stacked LoRA effectively mitigates inter-subject variability.
• Score: 0.34797121357690153
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Electroencephalography (EEG) decoding for brain-computer interfaces (BCIs) faces a major challenge: substantial inter-subject variability limits effective cross-subject generalization. Consequently, practical systems still rely largely on subject-specific models trained from scratch and requiring individual recalibration. EEG foundation models have recently emerged as a promising alternative; however, even large pretrained models cannot simply be used as fixed feature extractors and still require additional adaptation before they can be reliably applied to downstream tasks. In this work, we address this challenge through targeted adaptation strategies. Building on recent EEG foundation models such as REVE, LaBraM, and LUNA, we examine the impact of different low-rank adaptation strategies on motor imagery classification. We propose a framework that structurally decouples subject-invariant knowledge from subject-specific neural signatures: the low-rank update at each adapted layer is split into a Global adapter, trained jointly across all subjects, and Subject-Specific adapters, each absorbing individual variability. To assess the contribution of each path, we compare three adaptation strategies: (i) subject-specific LoRA (ii) global LoRA and (iii) stacked LoRA, combining both Global and Subject Specific adapters. Experiments on BCI Competition IV-2a, PhysioNet Motor Imagery, and the clinical Zuo2025 benchmark show that Stacked LoRA effectively mitigates inter-subject variability, achieving the best accuracy in the large majority of backbone and dataset combinations. Our analysis further reveals that the optimal balance between the global and subject-specific paths depends on the target population: a shared adapter is sufficient for large, diverse cohorts, whereas subject-specific adaptation is decisive under the high inter-session variability of clinical recordings.

Related papers

• Scaling Behavior Foundation Model for Humanoid Robots [77.31115920959364]
  Behavior Foundation Models (BFMs) have emerged as a promising solution to address challenges by leveraging large-scale behavioral data.<n>We show that our approach yields significant improvements in control fidelity and task generalization.<n>These results establish BFM as a principled and effective foundation for scalable and general-purpose humanoid control.
  arXiv  Detail & Related papers  (2026-07-16T16:08:27Z)
• LAtte: Hyperbolic Lorentz Attention for Cross-Subject EEG Classification [13.887781479252332]
  LAtte is a novel framework that integrates a Lorentz Attention Module with an InceptionTime-based encoder to enable robust and generalizable EEG classification.<n>First, we learn a shared baseline signal across all subjects using pretraining tasks to capture common underlying patterns.<n>Then, we utilize novel Lorentz low-rank adapters to learn subject-specific embeddings that model individual differences.<n>This allows us to learn a shared model that performs robustly across subjects, and can be subsequently finetuned for individual subjects or used to generalize to unseen subjects.
  arXiv  Detail & Related papers  (2026-03-11T15:27:39Z)
• Adapting Multimodal Foundation Models for Few-Shot Learning: A Comprehensive Study on Contrastive Captioners [1.2461503242570642]
  This paper presents a study on adapting the Contrastive Captioners (CoCa) visual backbone for few-shot image classification.<n>We identify an "augmentation divergence": while strong data augmentation degrades the performance of linear probing in low-shot settings, it is essential for stabilizing LoRA fine-tuning.<n>We also demonstrate that hybrid objectives incorporating Supervised Contrastive (SupCon) loss yield consistent performance improvements over standard Cross-Entropy.
  arXiv  Detail & Related papers  (2025-12-14T20:13:21Z)
• Rethinking Generalized BCIs: Benchmarking 340,000+ Unique Algorithmic Configurations for EEG Mental Command Decoding [0.0]
  We present a benchmark evaluating over 340,000+ unique combinations of spatial and nonlinear EEG classification.<n>Our findings highlight that no universal 'one-size-fits-all' method can optimally decode EEG motor imagery patterns across all users or datasets.
  arXiv  Detail & Related papers  (2025-12-02T17:56:46Z)
• TuckA: Hierarchical Compact Tensor Experts for Efficient Fine-Tuning [83.93651411533533]
  We introduce Tucker Adaptation (TuckA), a method with four key properties.<n>We develop an efficient batch-level routing mechanism, which reduces the router's parameter size by a factor of $L$.<n>Experiments on benchmarks in natural language understanding, image classification, and mathematical reasoning speak to the efficacy of TuckA.
  arXiv  Detail & Related papers  (2025-11-10T09:03:16Z)
• Mitigating Subject Dependency in EEG Decoding with Subject-Specific Low-Rank Adapters [0.0]
  We propose a drop-in replacement for standard linear or convolutional layers in any neural network architecture.<n>Our layer captures subject-specific variability by decomposing its weights into a shared, subject-invariant component.<n> Empirically, models equipped with our layer outperform both a shared-weight-only model and the average of individually trained subject-specific models.
  arXiv  Detail & Related papers  (2025-10-09T10:46:08Z)
• Explicit modelling of subject dependency in BCI decoding [12.17288254938554]
  Brain-Computer Interfaces (BCIs) suffer from high inter-subject variability and limited labeled data.<n>We present an end-to-end approach that explicitly models the subject dependency using lightweight convolutional neural networks (CNNs) conditioned on the subject's identity.
  arXiv  Detail & Related papers  (2025-09-27T10:51:42Z)
• MuSACo: Multimodal Subject-Specific Selection and Adaptation for Expression Recognition with Co-Training [52.99217736494484]
  We introduce MuSACo, a multi-modal subject-specific selection and adaptation method for personalized expression recognition.<n>This makes MuSACo relevant for affective computing applications in digital health, such as patient-specific assessment for stress or pain.<n>Our experimental results on challenging multimodal ER datasets: BioVid and StressID, show that MuSACo can outperform UDA (blending) and state-of-the-art MSDA methods.
  arXiv  Detail & Related papers  (2025-08-17T23:08:21Z)
• FACE: Few-shot Adapter with Cross-view Fusion for Cross-subject EEG Emotion Recognition [57.08108545219043]
  Cross-subject EEG emotion recognition is challenged by significant inter-subject variability and intricately entangled intra-subject variability.<n>Recent few-shot learning paradigms attempt to address these limitations but often encounter catastrophic overfitting during subject-specific adaptation with limited samples.<n>This article introduces the few-shot adapter with a cross-view fusion method called FACE for cross-subject EEG emotion recognition.
  arXiv  Detail & Related papers  (2025-03-24T03:16:52Z)
• Universal Semi-supervised Model Adaptation via Collaborative Consistency Training [92.52892510093037]
  We introduce a realistic and challenging domain adaptation problem called Universal Semi-supervised Model Adaptation (USMA) We propose a collaborative consistency training framework that regularizes the prediction consistency between two models. Experimental results demonstrate the effectiveness of our method on several benchmark datasets.
  arXiv  Detail & Related papers  (2023-07-07T08:19:40Z)
• Stacking Ensemble Learning in Deep Domain Adaptation for Ophthalmic Image Classification [61.656149405657246]
  Domain adaptation is effective in image classification tasks where obtaining sufficient label data is challenging. We propose a novel method, named SELDA, for stacking ensemble learning via extending three domain adaptation methods. The experimental results using Age-Related Eye Disease Study (AREDS) benchmark ophthalmic dataset demonstrate the effectiveness of the proposed model.
  arXiv  Detail & Related papers  (2022-09-27T14:19:00Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.