Robo Brain Learns Everything From The Internet

Robo Brain Learns Everything From The Internet

Robo-Brain-Learns-Everything-From-The-Internet

Last month, Cornell University researchers turned on Robo Brain, a large scale computational system that learns from publicly available internet resources, computer simulations, and real-life robot trials.

 

The robot is in the process of downloading about 1 billion images, 120,000 YouTube videos, and 100 million how-to documents and appliance manuals. The information will be translated and stored in a robot-friendly format that robots will be able to retrieve when needed.

 

“Our laptops and cellphones have access to all the information we want,” said Ashutosh Saxena, computer science assistant professor at Cornell University. “If a robot encounters a situation it hasn’t seen before it can query Robo Brain in the cloud.”

 

Robo Brain will act as a Wikipedia that robots can access to understand how we speak and how we see the world. For instance, if a robot sees a mug, it can learn from Robo Brain to recognize the mug is used to hold liquid, that it can be grasped by the handle and that it needs to be held upright when it is full, so as to avoid spillage, but can be tilted when it is empty, when it is being carried from the dishwasher to the cupboard.

 

The project is hosted at the official Robo Brain website, where users can assist by upvoting correct actions and objects with comments to the researchers. It is a system that will scour the internet for information and teach other robots. The system employs what the researchers call “structured deep learning” where information is stored in many layers of abstraction. For instance, if the robot sees an armchair, it knows it is a member of the class of chairs, and going up one level, chairs are furniture. Robo Brain knows that it is a furniture used for sitting, but that a human can also sit on a stool, a bench or the lawn.

How powerful is the database?

A robot’s computer brain stores what it has learned in a form mathematicians call a Markov model, which can be represented graphically as a set of points connected by lines (formally called nodes and edges). The nodes could represent objects, actions or parts of an image, and each one is assigned a probability – how much you can vary it and still be correct. In searching for knowledge, a robot’s brain makes its own chain and looks for one in the knowledge base that matches within those limits. “The Robo Brain will look like a gigantic, branching graph with abilities for multi-dimensional queries,” said Aditya Jami, a visiting researcher art Cornell, who designed the large-scale database for the brain. Perhaps something that looks like a chart of relationships between Facebook friends, but more on the scale of the Milky Way Galaxy.

 

Image courtesy of Wistan at deviantart.com


Did you remember to share this with your friends?