MIT and Harvard researchers demonstrate a better way for computers to see(web.mit.edu)
web.mit.edu
MIT and Harvard researchers demonstrate a better way for computers to see
http://web.mit.edu/press/2009/visual-systems.html
2 comments
Granted I just skimmed the article, but do they discuss the algorithm at all?? It seemed like they're just saying buzzwords, "parallel","gpu","vision is hard", etc.
Does anyone know if a new CV algorithm is involved in this?
Does anyone know if a new CV algorithm is involved in this?
They did link to the (open access) journal article: http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fj...
Based on the video they're tuning model parameters in bulk and screening for best performance at a defined task.
It's a really neat application of General Purpose GPU programming:
http://gpgpu.org/
http://en.wikipedia.org/wiki/GPGPU