More than a good eye: Robot uses arms, location and more to discover objects

May 06, 2013
Carnegie Mellon University researchers have shown that a two-armed mobile robot, called HERB, can continually discover and refine its understanding of objects by taking advantage of all of the information available, including the object's location, size, shape and even whether it can be lifted. Credit: Carnegie Mellon University

A robot can struggle to discover objects in its surroundings when it relies on computer vision alone. But by taking advantage of all of the information available to it—an object's location, size, shape and even whether it can be lifted—a robot can continually discover and refine its understanding of objects, say researchers at Carnegie Mellon University's Robotics Institute.

The Lifelong Robotic Discovery (LROD) process developed by the research team enabled a two-armed, to use color video, a Kinect depth camera and non- to discover more than 100 objects in a home-like laboratory, including items such as , plants and food items. Normally, the CMU researchers build digital models and images of objects and load them into the memory of HERB—the Home-Exploring Butler—so the robot can recognize objects that it needs to manipulate.

Virtually all roboticists do something similar to help their robots recognize objects. With the team's implementation of LROD, called HerbDisc, the robot now can discover these objects on its own. With more time and experience, HerbDisc gradually refines its models of the objects and begins to focus its attention on those that are most relevant to its goal—helping people accomplish tasks of daily living. Findings from the research study will be presented May 8 at the IEEE International Conference on Robotics and Automation in Karlsruhe, Germany.

The robot's ability to discover objects on its own sometimes takes even the researchers by surprise, said Siddhartha Srinivasa, associate professor of robotics and head of the Lab, where HERB is being developed. In one case, some students left the remains of lunch—a pineapple and a bag of bagels—in the lab when they went home for the evening. The next morning, they returned to find that HERB had built digital models of both the pineapple and the bag and had figured out how it could pick up each one.

"We didn't even know that these objects existed, but HERB did," said Srinivasa, who jointly supervised the research with Martial Hebert, professor of robotics. "That was pretty fascinating."

Discovering and understanding objects in places filled with hundreds or thousands of things will be a crucial capability once robots begin working in the home and expanding their role in the workplace. Manually loading of every object of possible relevance simply isn't feasible, Srinivasa said. "You can't expect Grandma to do all this," he added.

Object recognition has long been a challenging area of inquiry for researchers. Recognizing objects based on vision alone quickly becomes an intractable computational problem in a cluttered environment, Srinivasa said. But humans don't rely on sight alone to understand objects; babies will squeeze a rubber ducky, beat it against the tub, dunk it—even stick it in their mouth. Robots, too, have a lot of "" about their environment that they can use to discover objects.

Taking advantage of all of HERB's senses required a research team with complementary expertise—Srinivasa's insights on robotic manipulation and Hebert's in-depth knowledge of computer vision. Alvaro Collet, a robotics Ph.D. student they co-advised, led the development of HerbDisc. Collet is now a scientist at Microsoft.

Depth measurements from HERB's Kinect sensors proved to be particularly important, Hebert said, providing three-dimensional shape data that is highly discriminative for household items.

Other domain knowledge available to HERB includes location—whether something is on a table, on the floor or in a cupboard. The robot can see whether a potential object moves on its own, or is moveable at all. It can note whether something is in a particular place at a particular time. And it can use its arms to see if it can lift the object—the ultimate test of its "objectness."

"The first time HERB looks at the video, everything 'lights up' as a possible object," Srinivasa said. But as the robot uses its domain knowledge, it becomes clearer what is and isn't an object. The team found that adding domain knowledge to the video input almost tripled the number of objects HERB could discover and reduced computer processing time by a factor of 190. A HERB's-eye view of objects is available on YouTube.

HERB's definition of an object—something it can lift—is oriented toward its function as an assistive device for people, doing things such as fetching items or microwaving meals. "It's a very natural, robot-driven process," Srinivasa said. "As capabilities and situations change, different things become important." For instance, HERB can't yet pick up a sheet of paper, so it ignores paper. But once HERB has hands capable of manipulating paper, it will learn to recognize sheets of paper as objects.

Though not yet implemented, HERB and other robots could use the Internet to create an even richer understanding of objects. Earlier work by Srinivasa showed that robots can use crowdsourcing via Amazon Mechanical Turk to help understand objects. Likewise, a robot might access image sites, such as RoboEarth, ImageNet or 3D Warehouse, to find the name of an object, or to get images of parts of the object it can't see.

Explore further: New RFID technology helps robots find household objects

Related Stories

New robot butler "HERB" can microwave your dinner (w/ Video)

May 03, 2012

(Phys.org) -- One of the big disappointments of the computer age is the distinct lack of robots in our everyday lives. For years we’ve all been teased by the possibilities of robots in SciFi movies and television shows, ...

Robots learn to pick up oddly shaped objects

May 09, 2012

(Phys.org) -- When Cornell engineers developed a new type of robot hand that could pick up oddly shaped objects it presented a challenge: It was easy for a human operator to choose the best place to take h ...

'Hallucinating' robots arrange objects for human use

Jun 18, 2012

(Phys.org) -- If you hire a robot to help you move into your new apartment, you won't have to send out for pizza. But you will have to give the robot a system for figuring out where things go. The best approach, ...

Recommended for you

New RFID technology helps robots find household objects

Sep 22, 2014

Mobile robots could be much more useful in homes, if they could locate people, places and objects. Today's robots usually see the world with cameras and lasers, which have difficulty reliably recognizing ...

Victoria team defend title with speedy robot

Sep 22, 2014

A team from Victoria's School of Engineering and Computer Science, led by Robby Lopez, beat 15 other teams from Australian and New Zealand universities to take top honours in the 2013 competition with its ...

User comments : 0