News for hackers. Stuff that matters.
Re:UC Berkeley Scientists Translate Brainwaves into Imagery by antimora on Thursday September 22, 2011 @11:05PM (#3028677)
The article is misleading. There is not deconstruction or decoding of brain waves happening, it's simply correlating stimuli with prior learned/trained images. If the model was trained on images of dogs and cats, the "reconstructed" images will be terms of dogs and cats, which is the basic limitation of the process.

Source: "In practice fitting the encoding model to each voxel is a straightforward regression problem. " (http://gallantlab.org/)

Re:UC Berkeley Scientists Translate Brainwaves into Imagery by glenstein on Friday September 23, 2011 @09:26AM (#3029877)
The reconstructed image would also be in terms of the abstract features of dogs and cats: the shades of color, contours of their bodies, their position on the screen. And the abstract features could be recombined into an average image that's completely unlike any dog or cat but resembles what a person is looking at.
Re:UC Berkeley Scientists Translate Brainwaves into Imagery by bugsbunnyak on Friday September 23, 2011 @12:30AM (#3028830)
Another group does it directly from fMRI: http://www.ncbi.nlm.nih.gov/pubmed/20460157

Among the caveats: really high-T magnet, not feasible for general use; the visual stimulus was actually a warped version that would be reproduced as the nicely shaped activation upon retinotopic projection (it's a log-polar mapping)... Still really cool.

(ps: no-paywall version here: surfer.nmr.mgh.harvard.edu)