ALFRED Challenge
Leaderboard is now live for submissions to the ALFRED challenge! Humans have a success rate of 91% on unseen environments, but our baseline model has a 0.4% success rate. š¢
Challenge Winner: After 17 submissions, we saw a 10x increase in success rate to 4.5% by the challenge winners: Van-Quang Nguyen and Takayuki Okatani of Tohoku University!
Can you do even better? Code, precomputed features, and AI2Thor simulator are all available for a quick start on GitHub