AI from Carnegie Mellon Specializes in Images
March 6, 2014
Here’s yet another personified AI attempting to mimic the human brain, and this one focuses on processing pictures. TechRadar invites us to “Meet NEIL, the Computer that Thinks Like You Do.” A team from Carnegie Mellon has developed NEIL (Never Ending Image Learner) specifically to interpret images and make connections between them. Writer Dean Evans reports:
According to Xinlei Chen, a PHd student who works with NEIL, the software “uses a semi-supervised learning algorithm that jointly discovers common sense relationships – e.g ‘Corolla is a kind of/looks similar to Car’, ‘Wheel is part of Car’ – and labels instances of the given visual categories… The input is a large collection of images and the desired output is extracting significant or interesting patterns in visual data – e.g. car is detected frequently in raceways. These patterns help us to extract common sense relationships.
As the ‘never ending’ part of its name suggests, NEIL is being run continuously, and it works by plundering Google Image Search data to amass a library of objects, scenes and attributes. The current array of information includes everything from aircraft carriers to zebras, basilicas to hospitals, speckled textures to distinctive tartan patterns.
Of course, NEIL is not perfect; it has incorrectly linked windmills with helicopters and radiators with accordions, for example. Still, its success rate was pegged at 79 percent in a random sample. See the article for more information on how the system works.
NEIL might be considered the little brother to NELL, the Never Ending Language Learner, also built by researchers at Carnegie Mellon. NELL’s specialty is “to ‘read the web’ and to extract a set of true, structured facts from the pages that it analyses.” NELL has been at it since 2010, and has come to over two million conclusions. Will the University continue adding to the family?
Cynthia Murrell, March 06, 2014
Sponsored by ArnoldIT.com, developer of Augmentext