MIT researchers have developed a machine-learning model that highlights objects in an image as their described in audio captions in real-time. The work is out of the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL).

Sourced through Scoop.it from: news.mit.edu