Mind-Video

Mind-Video is an advanced research tool that reconstructs continuous video representations of visual experiences directly from non-invasive human brain activity (fMRI data). Developed by the Mind-X research group, Mind-Video uses a two-module pipeline featuring an fMRI encoder and an augmented stable diffusion model to generate semantically meaningful videos from brain signals. This platform is ideal for neuroscience researchers, brain-computer interface developers, and cognitive scientists seeking to bridge the gap between neural data and visual reconstruction in both academic and applied research contexts.