Published March 20, 2026
| Version v1
Preprint
Open
Zero_Shot_Motor_Imagery_BCI_Paper
Authors/Creators
Description
Zero-Shot Learning for Motor Imagery-based Brain-Computer Interfaces (BCI).
One of the primary bottlenecks in current BCI technology is the extensive calibration period required for each new user. This paper explores a zero-shot approach, aiming to bypass subject-specific training by leveraging cross-subject data and advanced machine learning architectures.
Key highlights of the work:
- Calibration-free BCI: Reducing the barrier to entry for new users by eliminating lengthy setup phases.
- EEG Signal Processing: Implementing robust feature extraction to handle non-stationarity in brain signals.
- Generalization: Testing the model's ability to interpret motor imagery (imagined movements) on previously unseen subjects.
Files
Zero_Shot_Motor_Imagery_BCI_Paper.pdf
Additional details
Related works
- Cites
- Preprint: 10.5281/zenodo.17506760 (DOI)
Software
- Programming language
- Python console , Python
- Development Status
- Active