Google AI researchers used 2,000 mannequin challenge videos from YouTube to train AI that predicts depth in videos with people and objects in motion.Read More

>> Source: VentureBeat