ReCIL uses Euclidean alignment to reduce inter-subject EEG distribution shift, global-local replay to preserve old-task knowledge, and dimensionality reduction for reliable similarity computation; experiments on three public MI datasets showed a good balance between plasticity and stability, and the authors report this is the first study on cross-subject class incremental learning for MI classification. PMID 42560905, DOI 10.1109/tbme.2026.3721154
ReCIL: Rehearsal-Based Class Incremental Learning for Cross-Subject Motor Imagery Classification
Summary
Researchers propose ReCIL, a rehearsal-based class incremental learning method for cross-subject motor imagery classification that lets a model learn new MI classes sequentially without retraining from scratch. Using Euclidean alignment to reduce cross-subject EEG distribution shift and global-local replay to preserve earlier-task knowledge, ReCIL achieved a good balance between plasticity and stability across three public MI datasets. The authors report it as the first study of cross-subject class incremental learning for MI classification.
Why it matters
The method targets a practical bottleneck: existing MI decoders must re-collect data or be retrained whenever a new class is added, so incremental learning points to brain-computer interfaces whose command vocabularies can grow without that recurring cost.
Compiled by BCIwiki from public sources
Sources · 1
pubmed.ncbi.nlm.nih.gov 2026-08-06