Home Household chores Scientists teach AI to do household chores

Scientists teach AI to do household chores


Researchers from the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT and the University of Toronto teach AI how to perform household chores using a virtual simulator inspired by The Sims. While household chores are dreary if not particularly complex for humans, for the AI ​​making a cup of coffee or turning on the toaster is a bit more complex. Watching TV from the sofa, for example, requires an AI to break the task down into specific steps, such as: “walk to TV”, “turn on TV”, “walk to sofa” and “sit on the couch”. .”

To complete the chores, the researchers collected common verbal descriptions of household chores and then translated them into robot-friendly code. Next, they introduced AI home assistants in a VirtualHome 3D stimulator that looks a lot like The Sims. Not only did the virtual robot agents manage to make coffee, turn on the toaster and relax on the couch, but the researchers created a database of household chores described in natural language, which could help systems like Amazon Alexa to perform more complex tasks.

This content is imported from YouTube. You may be able to find the same content in another format, or you may be able to find more information, on their website.

In the future, assistant robots might even learn by watching Youtube videos instead of stimulation videos. The researchers also want to implement a rewards system that gives the AI ​​positive feedback when it performs tasks correctly. In the future, robot assistants may be able to learn and perform complex, multi-step tasks simply by watching a human perform them.

“You can imagine an environment where robots help with household chores and can possibly anticipate personalized wants and needs, or impending action,” said PhD student Xavier Puig, lead author of the research. “It could be particularly useful as an assistive technology for the elderly or those with reduced mobility.”

The team’s AI agent has already learned to perform 1,000 distinct sets of actions in eight different scenes, including a living room, kitchen, dining room, bedroom, and home office.

(Going through MIT CSAIL)

This content is created and maintained by a third party, and uploaded to this page to help users provide their email addresses. You may be able to find more information about this and similar content on piano.io