Researchers have developed a novel framework for motor imagery classification using a hybrid electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) approach. This framework combines Cross-Subject Class Mixup (CSCM) with Selective Attention to improve generalization across different subjects. CSCM generates synthetic training data by mixing trials from various subjects, while Selective Attention effectively models temporal signal characteristics and fuses the two modalities. In leave-one-subject-out evaluations on a dataset of 28 subjects, the model achieved a mean balanced accuracy of 76.85% and a macro-averaged F1 score of 76.58%, outperforming five baseline methods. AI
IMPACT This research advances brain-computer interface capabilities by improving the accuracy and generalizability of motor imagery classification.
RANK_REASON The cluster contains a research paper detailing a new methodology for motor imagery classification. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Hugging Face Daily Papers →
- brain–computer interface
- Cross-Subject Class Mixup
- electroencephalography
- fNIRS
- Hugging Face
- motor imagery
- selective attention
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →