Age-Old 3D Computer Vision Problem Solved?
Posted Tuesday, June 13th, 2006, at 6:40 pm Eastern by Mark Wallace
Well, as old as the computer age, anyway. Teaching machines to correctly discern objects in three-dimensional scenes has long been a sticking point in artificial intelligence and robotics. Now, according to a UPI story, “Carnegie Mellon University scientists say it is now possible for computers to learn how to discern the geometric context of natural scenes, which, previously, has been a major roadblock for computer vision.” Said scientists will apparently be unveiling their creation at the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, which starts this weekend. Intriguing. And these are the real deep geeks. Maybe I’ll try to go. [Thanks to Glitchy for the link.]
Comments are closed. Trackbacks are closed.
Feeds
finally the robots will be able to get around town easier