Indian-origin researchers have created 'Robo Brain', a large-scale computational system that learns from internet, to teach other robots how the world works and how humans behave.

Currently, Robo Brain is processing about 120,000 YouTube videos and a billion images from internet. According to researchers, the information is being translated and stored in the format that robots can understand. When needed, the robots can draw the information.

Ashutosh Saxena, lead researcher and assistant professor of computer science at Cornell University in New York, said, "Our laptops and cellphones have access to all the information we want. If a robot encounters a situation it has not seen before, it can query 'Robo Brain' in the cloud".

After understanding how the world works, the robots can help humans in homes and offices. By connecting videos and pictures, Robo Brain will help a robot perform any action. With the help of Robo Brain, a robot can learn about existing things. For example, Robo Brain will not only tell a robot about a coffee mug, but also tell that robot has to grasp it by handle as the liquid inside the mug can be poured out.

The researchers have explained about Robo Brain at the "2014 Robotics: Science and Systems Conference" in Berkeley. In the conference, they have described where information is stored in the Robo Brain.
Saxena said that to recognize objects and their working, Robo Brain will connect videos and images with text.

A visiting researcher at Cornella and designer of the large-scale database for the brain, Aditya Jami, said that Robo Brain has the ability for multi-dimensional queries.

Companies like Google and Microsoft are also working on deep learning technologies. Recently, Microsoft has shown its Project Adam where Microsoft's voice assistant Cortana identifies objects.