Researchers at Cornell and Stanford universities have developed Watch-bot, a robot that can independently learn a user's household activity patterns to provide helpful reminders.
Watch-bot consists of a Microsoft Kinect three-dimensional sensor, a camera that can pan and tilt, a laptop, and a laser pointer.
During development, the robot was set up in a kitchen and an office, and watched people go about their daily routines for a week. The system collected 458 videos of normal human activity, which were then annotated with 21 different actions and 23 types of objects.
In 222 of the videos, someone intentionally forgot to do something, and Watch-bot was able to identify the forgotten action about 60% of the time. Tasks that Watch-bot provided reminders for included putting a book back on a shelf after reading it, turning a monitor off when a computer is not being used, putting milk back into the fridge, and getting food from the microwave.
The system uses probabilistic learning models capable of detecting patterns and relationships directly from the camera and Kinect data.
Watch-bot is not meant to be a functional system, but a proof of concept of the underlying technology, which can be transferred to a wide variety of robots.
From IEEE Spectrum
View Full Article
Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA
No entries found