In the Media

Engadget reporter Christine Fisher writes that MIT researchers have developed a new technique that improves the speed and performance of video recognition models. Fisher writes that the new method “reduces the size of video-recognition models, speeds up training and could improve performance on mobile devices.”