SumabilitySum It Up - Your AI for Content Curation. 🤖

Feedback

Bug Report
Feature Request

Twitter page

Topics

AI 🤖Business 💼Cryptocurrency 💰Economy 🏛️Politics 🥸Stock Markets 🤑Technology 🛠️

Posted 9/17/2023, 11:00:00 AM

NYU Professor Named Top Young Innovator for Teaching Robots with Trial and Error

MIT Technology Review released its annual list of top innovators under 35, including NYU professor Lerrel Pinto.
Pinto teaches robots household tasks by letting them fail and learn.
He set up a model home with toy appliances for the robots to practice with.
The technique is called reinforcement learning - the robot tries something, judges its performance, and tries again.
Pinto published a paper "Teach a Robot to FISH" with examples of how tasks were trained.

hackaday.com

Relevant topic timeline:

8/22/2023

A deep learning technique to improve how robots grasp objects

Researchers at the University of Bonn's Autonomous Intelligent Systems research group have developed a deep reinforcement learning pipeline to enhance a robotic arm's ability to manipulate objects in ways that better support practical use, potentially contributing to the development of more effective robotic assistants.

9/19/2023

Toyota Uses Kindergarten to Teach Robots Skills By Showing, Not Coding

Toyota Research Institute (TRI) has used generative AI to teach robots how to perform individual tasks needed to make breakfast by giving them a sense of touch and plugging them into an AI model, enabling them to "learn" and carry out tasks in a matter of hours.

9/20/2023

Toyota's New AI Allows Robots to Master Complex Kitchen Tasks Just by Watching Humans

Toyota has developed a new learning approach called Diffusion Policy, in partnership with Columbia Engineering and MIT, which enables robotic AIs to learn complex physical tasks by observing human demonstrations, allowing them to quickly acquire new skills and perform a wide variety of tasks with flexibility and adaptability.

Home
Technology 🛠️