Decoding visual images directly from brain scans

This is really remarkable. Scientists at UC Berkeley used functional magnetic resonance imaging (fMRI) to scan the brains of members of their own team while the subjects watched movie trailers, then scans were run through computer analysis to try to reconstruct them by combining a large selection of YouTube videos not including the clips in question. The results were far from perfect, but way better than I'd have expected. Previous work was with black-and-white still pictures; this time it's moving video. Below you can see samples of the original video together with the attempted reconstruction from the brain scans:

For more, see Berkeley's press release.


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>