HMT-Net: A Hybrid Mamba–Transformer Network for Motor Imagery EEG Decoding
Mathematics · 15.06.2026 · 14 (12) · 2149
DOI: 10.3390/math14122149 ↗Kurz gefasst
Fachartikel in der Zeitschrift Mathematics, online erschienen am 15.06.2026. Verfasst von Tingting Zhang und 6 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Key Research and Development Program of China, Natural Science Foundation of Fujian Province of China.
Abstract (Originalsprache)
Electroencephalography (EEG) is widely used in brain-computer interfaces (BCIs) for decoding motor imagery (MI) signals. However, existing methods remain limited in extracting multi-scale local spatiotemporal features and effectively integrating them with global feature information, leaving room for further improvement in classification accuracy. To address this issue, we propose HMT-Net, a hybrid architecture that integrates multi-scale convolution, the Mamba state-space model, and a self-attention mechanism. The model consists of a shallow feature embedding (SFE) module for spatiotemporal feature extraction, a multi-scale local feature extractor (MSLFE), and a Mamba–transformer global feature encoder (MTGFE). Specifically, the MSLFE employs dual-branch convolutions and channel attention to achieve adaptive multi-scale perception, while the MTGFE combines Mamba’s linear sequence modeling capability with multi-head attention to efficiently capture global dependencies. Unlike conventional Mamba or transformer EEG models, HMT-Net couples linear state-space modeling with global pairwise attention, avoiding the representational limits inherent in each individual architecture. Experiments on the BCI-IV-2a, BCI-IV-2b, and HGD datasets show that HMT-Net achieves subject-dependent accuracies of 84.07%, 89.60%, and 96.02%, respectively, outperforming EEGNet, FBCNet, EEGConformer, and ATCNet by 11.65%, 5.02%, 5.13%, and 6.60%, respectively, on BCI-IV-2a. Furthermore, HMT-Net achieves the best accuracy in subject-independent experiments, demonstrating strong generalization capability. Ablation studies and visualizations further validate the effectiveness and interpretability of the proposed model.
Förderung
- National Key Research and Development Program of China
- Natural Science Foundation of Fujian Province of China
Zitieren
APA
Zhang, T., Liao, H., Mu, Y., Han, J., Li, N., Hu, G., & Kong, X. (2026). HMT-Net: A Hybrid Mamba–Transformer Network for Motor Imagery EEG Decoding. Mathematics, 14(12), 2149. https://doi.org/10.3390/math14122149
Vancouver
Zhang T, Liao H, Mu Y, Han J, Li N, Hu G, et al. HMT-Net: A Hybrid Mamba–Transformer Network for Motor Imagery EEG Decoding. Mathematics. 2026;14(12):2149. doi:10.3390/math14122149
BibTeX
@article{Zhang2026,
author = {Zhang, Tingting and Liao, Haorong and Mu, Yiming and Han, Junfeng and Li, Nan and Hu, Guoyu and Kong, Xiangzeng},
title = {HMT-Net: A Hybrid Mamba–Transformer Network for Motor Imagery EEG Decoding},
journal = {Mathematics},
year = {2026},
volume = {14},
number = {12},
pages = {2149},
doi = {10.3390/math14122149}
}
EUVIMED