ニュース&特集
Microsoft's CaptionBot uses computer vision and natural language processing to describe any photo, just as a human would.
ニュース | TechCrunch
Microsoft demos next-generation image-captioning Captionbot
Microsoft created Captionbot.ai, which is a tremendously addictive (and science-fiction-grade awesome).
ニュース | Communications of the ACM
Seeing More Clearly (ACM .pdf)
Communications of the ACM interviewed Fei-Fei Li, Rob Fergus, Richard Zemel and Xiaodong He on recent progress in computer vision and language processing, interview highlighted in "Seeing More Clearly" in the January 2016 issue of CACM.
ニュース | The Guardian
Microsoft has made a computer that can recognise a dog in a bike basket
The system uses natural language processing, which breaks down a scene into simple language that a human would use (e.g. what is in the left side behind me?).
Microsoft Research and a team at Carnegie Mellon University have developed a system that can train machines to examine an image and seek to answer questions the way a human might ask them.