Physiological SIgnals as Input for GenAI

Created using ChatSlide

This research presents an exploration of EEG-based generative systems, highlighting the advantages and challenges of EEG over fMRI, such as noise and spatial resolution. It introduces the GWIT framework for EEG-to-image translation, detailing its pipeline, including EEG conditioning and ControlNet adapters. The study examines its implications for advancements in Brain-Computer Interface applications.