Communications of the ACM
Experts Establish 'North Star' for Domestic Robotics Field
In creating the BEHAVIOR benchmark, the researchers have established a visual reference point by which to gauge the success of future artificial intelligence solutions, which might also be used to develop and train robotic assistants in virtual environmen
Credit: Miriam Doerr
Stanford University researchers have created a catalogue of the physical and intellectual details of 100 everyday household tasks, like washing dishes and cleaning countertops, to set benchmarks for domestic robots.
The Benchmark for Everyday Household Activities in Virtual, Interactive, and Ecological Environments (BEHAVIOR) details the implementation of those tasks in multiple simulated homes, features a three-dimensional simulator with a virtual reality (VR) interface, and includes success metrics based on humans performing the same tasks in VR.
Stanford's Jiajun Wu said, "While any one of those tasks is already highly complex in its own right, imagine the challenge of creating a single robot that can do all of these things. Creating these benchmarks now, before the field has evolved too far, will help to set up potential common goals for the community."
From Stanford University
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA
No entries found