ZUNA EEG is a BCI model that reliably reconstructs missing or corrupted EEG channels and performs spatial upsampling to increase the effective resolution of EEG signals.
MODALITIES
Electroencephalography (EEG)
Brain Computer Interface (BCI)
ARCHITECTURE
380M Parameters Diffusion Autoencoder
FEATURES
Montage-Agnostic Scaling
Works across any EEG channel configuration.
Hardware-efficient gains
Improves signal quality and downstream performance.
Cross-dataset generalization
Generalizes across acquisition systems.
Real-world EEG ready
Built for noisy, incomplete, and motion-corrupted data.
