The robot takeover is officially underway. A computational system called Robo Brain teaches itself by using publicly available Internet resources. It is currently downloading and processing about a billion images, 120,000 YouTube videos, and 100 million how-to documents and manuals. The information will be translated and stored in a format that will allow robots to access and learn from.
For robots to truly become part of the workforce in offices and factories or to become assistants at home, they need to be taught about human behavior. Researchers have been teaching them things one-by-one, such as how to put away dishes or find your keys. With Robo Brain, this information will be available all at once.
"Our laptops and cell phones have access to all the information we want. If a robot encounters a situation it hasn't seen before it can query Robo Brain in the cloud," Ashutosh Saxena, an assistant professor of computer science at Cornell University, told PhysOrg.
The Robo Brain will work by identifying objects in images by connecting the images to videos and text. This will allow it to recognize objects and learn how they are used with human language and behavior.
So, if a robot sees a coffee mug, it can use the Robo Brain to identify it as a mug and learn that liquids can be poured into it, can be held by holding the handle and should be carried upright when full.
"The Robo Brain will look like a gigantic, branching graph with abilities for multi-dimensional queries," explains Aditya Jami, a visiting researcher at Cornell who designed the brain's database.
Learn more about the Robo Brain in the video below.