EEG-based fatigue detection requires modeling spectral information across bands and its temporal variation. Recent methods combine multi-band modeling with adaptive temporal processing, yet how spectral structure across bands should influence the contribution of each temporal interval remains insufficiently explored. We propose LBSNet, a lightweight network that uses prototype-based brain-state modeling to link local spectral structure with temporal aggregation. LBSNet extracts compact multi-band features through band-specific temporal convolution, shared spatial projections, and cross-band recalibration. Within each window, relative spectral activity is encoded as a normalized descriptor and softly matched to learnable spectral prototypes. Prototype assignments and learned prototype importance jointly determine temporal weights for power aggregation, allowing the spectral pattern of each window to guide its contribution to the prediction. Subject-independent experiments on SEED-VIG and SADT show that LBSNet outperforms seven baselines while maintaining low computational cost. Ablation and visualization analyses support prototype-guided temporal weighting, while the learned prototypes reveal a gradual redistribution of relative cross-band spectral states with increasing fatigue. The code is publicly available at https://github.com/zy0929/LBSNet.
Figures & tables
Figure 1: Overview of LBSNet, where learnable spectral prototypes represent local relative spectral patterns and guide temporal aggregation.
Method
SEED-VIG
SADT
Accuracy
F1-score
Accuracy
F1-score
EEGNet
56.92 ± 1.63
53.71 ± 1.16
67.42 ± 2.76
66.62 ± 3.10
TCNet
57.01 ± 2.72
54.26 ± 2.52
68.75 ± 1.66
68.11 ± 1.89
Conformer
58.08 ± 2.03
55.09 ± 2.47
70.38 ± 1.66
69.72 ± 2.02
Deformer
58.53 ± 2.72
55.21 ± 2.40
70.61 ± 2.43
69.91 ± 3.07
BiConformer
58.70 ± 1.76
55.92 ± 1.58
70.71 ± 1.74
70.18 ± 1.88
Table 1: Performance comparison (%) on SEED-VIG and SADT.
Method
SEED-VIG
SADT
Accuracy
F1-score
Accuracy
F1-score
Uniform Pooling
60.25 ± 1.49
58.61 ± 2.66
71.96 ± 2.24
72.49 ± 2.48
MLP Attention
60.63 ± 2.37
58.01 ± 2.46
72.55 ± 0.96
71.99 ± 1.00
w/o CBR
61.49 ± 1.42
58.07 ± 2.23
72.23 ± 1.37
72.68 ± 1.54
LBSNet
62.73 ± 1.82
61.51 ± 1.39
73.74 ± 1.38
73.30 ± 1.43
Table 2: Ablation results (%) on SEED-VIG and SADT.
Figure 2: Comparison of model accuracy and computational efficiency on SADT. Each bubble corresponds to a model. The x-axis shows MACs (M) on a logarithmic scale, while the y-axis reports ACC (%). Bubble area reflects the logarithm of the number of trainable parameters, and the labels give the corresponding parameter counts.
Figure 3: t-SNE visualization of representations produced by temporal aggregation on SADT. (a) Shuffled weights. (b) Learned weights.
Figure 4: Analysis of learned spectral prototypes on SEED-VIG. (a) Relative spectral patterns of the learned prototypes across the five frequency bands. (b) Prototype proportions for alert and fatigue samples. (c) Changes in prototype proportions across PERCLOS values. The dashed line marks the 0.35 threshold between alert and fatigue, and the histogram shows the sample distribution.
Electroencephalography (EEG) studies brain-generated electrical activity to understand cognition and support diagnosis, monitoring, and brain-computer interfaces.
Cross-subject EEG stress detection remains challenging because discriminative stress-related patterns are both subject-dependent and frequency-specific. Conventional Riemannian methods model spatial covariance mainly in the time domain, overlooking neural oscillations that are critical for high-level cognitive state decoding, while standard temporal tokenization often fragments inter-slice temporal coherence. To address these limitations, we propose \method{}, an Intra-Inter Riemannian Manifold Attention Network for EEG-based stress detection. \method{} constructs spatial covariance matrices independently at each frequency point and maps them to the SPD tangent space, preserving channel-wise geometry together with frequency-specific discriminative cues. It further introduces frequency cluster aggregation to select informative spectral components and reduce redundancy by forming compact, data-driven frequency clusters aligned with EEG rhythms. Finally, an intra-inter slice attention module adaptively integrates local slice-level spectral dynamics and global temporal context across EEG sequences. Experiments on three datasets show that \method{} consistently outperforms five state-of-the-art baselines, achieving up to 82.78% balanced accuracy while remaining efficient with only 1.60M parameters and 31.95M FLOPs.
Cheng He, Kunyu Peng, Shangen Han +3
Laboratory of Neural Computing and Intelligent Perception, College of Information Engineering, Capital Normal University, Beijing, China · Karlsruhe Institute of Technology, Germany · School of Psychology, Capital Normal University, Beijing, China
Electroencephalography (EEG) is a widely used tool for studying brain function, with applications in clinical neuroscience, diagnosis, and brain-computer interfaces (BCIs). Recent EEG foundation models trained on large unlabeled corpora aim to learn transferable representations, but their effectiveness remains unclear; reported improvements over smaller task-specific models are often modest, sensitive to downstream adaptation and fine-tuning strategies, and limited under linear probing. We hypothesize that one contributing factor is the reliance on signal reconstruction as the primary self-supervised learning (SSL) objective, which biases representations toward high-variance artifacts rather than task-relevant neural structure. To address this limitation, we explore an SSL paradigm based on Joint Embedding Predictive Architectures (JEPA), which learn by predicting latent representations instead of reconstructing raw signals. We introduce Laya, the first EEG foundation model based on LeJEPA. We show that latent prediction yields representations that encode semantic structure in EEG: Laya embeddings track clinically meaningful state changes such as seizure onset, are resilient to noise, and achieve the strongest mean clinical accuracy under frozen linear probing, with particular gains on tasks where relevant neural patterns are subtle and easily obscured by artifacts. Controlled ablations against matched MAE variants confirm that the choice of pretraining objective, rather than architecture or data, is the primary driver of these gains.
Saarang Panchavati, Uddhav Panchavati, Hiroki Nariai +2
Department of Medical Informatics UCLA Los Angeles, USA · Department of Cognitive Science UCSD San Diego, USA · David Geffen School of Medicine UCLA Mattel Children’s Hospital Los Angeles, USA
Electroencephalograph (EEG) timeseries signals are characterized by significant noise and coarse spatial resolution, which complicates the classification of neurodegenerative diseases. Even SOTA deep learning architectures struggle to distinguish between healthy controls and diseased subjects, or between different disease types, due to high intergroup similarity. In this paper, we show that a spectrally selective approach to feature construction enhances class separability. By isolating signal strengths within the primary brainwave bands, we transform high dimensional raw data into high value spectral features. Our results demonstrate that in small datasets a) features derived from frequency and time frequency domain allow traditional machine learning models to match or exceed the performance of SOTA deep learning models, b) Attention mechanism is unable to distill the stable feature signatures that characterize healthy neural activity in both resting and task EEGs, and c) the limitations of attention based models in finding relevant spectral features appear to be robust in that providing frequency selective time domain input do not appreciably improve their performance. We validate our methodology across three open source resting EEG datasets and one task EEG dataset, providing robust empirical evidence for our claims.