Robots Learn By Watching Videos

Are you among those who often dream of a day when a robot will do all the everyday household chores for you? A team of researchers from Carnegie Mellon University (CMU) has figured out how to turn your dream into reality.

In their latest study, they proposed a model that allowed them to train robots to do household tasks by showing them videos of people doing ordinary activities in their homes, like picking up the phone, opening a drawer, etc.

So far, scientists have been training robots by physically showing them how a task is done or training them for weeks in a simulated environment. Both these methods take a lot of time and resources and often fail.

The CMU team claims that their proposed model, Visual-Robotics Bridge (VRB), how can make a robot learn a task in just 25 minutes, and that too without involving any humans or simulated environment.

This work could drastically improve the way robots are trained and “could enable robots to learn from the vast amount of internet and YouTube videos available," said Shikhar Bahl, one of the study authors and a Ph.D. student at CMU’s School of Computer Science.

Robots have learned to watch and learn
VRB is an advanced version of WHIRL (In-the-Wild Human Imitating Robot Learning), a model that researchers used previously to train robots.  READ MORE...