Training machines to learn more like humans do : vimarsana.c

MIT researchers discovered that a specific training technique can enable certain types of computer vision models to learn more stable, predictable visual representations, which are more similar to those humans learn using a biological property known as perceptual straightening.

Related Keywords

Williamt Freeman , Ruth Rosenholtz , Anne Harrington , Vasha Dutell , Mark Hamilton , Ayush Tewari , Department Of Brain , Dana Farber Cancer Institute , Artificial Intelligence Laboratory , National Science Foundation , International Conference On Learning Representations , Harvard Medical School , Us Air Force Artificial Intelligence Accelerator , Toyota Research Institute , York University , Us Air Force Research Laboratory , Cognitive Sciences , Department Of Electrical Engineering , Electrical Engineering , Computer Science , Simon Stent , Woven Planet , Gerd Perkins Professor , International Conference , New York University , Bill Lotter , Force Research Laboratory , Force Artificial Intelligence , Erceptual Straightening , Computer Vision , Machine Learning , Adversarial Training ,