title

Reward Isn’t Free: Supervising Robot Learning with Language and Video from the Web | SAIL Blog

title

Reward Isn't Free: Supervising Robot Learning with Language and Video from the Web | The Stanford AI Lab Blog

charset

utf-8

viewport

width=device-width, initial-scale=1, maximum-scale=1

generator

Jekyll v3.9.0

og:title

Reward Isn’t Free: Supervising Robot Learning with Language and Video from the Web

og:locale

en_US

og:description

This work was conducted as part of SAIL and CRFM.

og:url

http://ai.stanford.edu/blog/reward-isnt-free/

og:site_name

SAIL Blog

twitter:card

summary

twitter:title

Reward Isn't Free: Supervising Robot Learning with Language and Video from the Web

twitter:description

Where do the rewards for robotic reinforcement learning come from? In this blog post we study how using crowdsourced language annotations and videos of humans, we can learn reward functions in a scalable way and enable them to generalize more broadly.

twitter:creator

@StanfordAILab

twitter:card

summary_large_image

alternate

http://ai.stanford.edu/blog/feed.xml

canonical

http://ai.stanford.edu/blog/reward-isnt-free/

canonical

http://ai.stanford.edu/blog/reward-isnt-free/

icon

/blog/assets/img/favicon-32x32.png

icon

/blog/assets/img/favicon-16x16.png