"Vision-and-Dialog Navigation" will appear at CoRL'19 in Osaka later this month! Our leaderboard for the navigation challenge is up at https://t.co/l7gHGHNyNp through #EvalAI thanks to @mmurz.
@project_cloudcv
Check out our new preprint, "Vision-and-Dialog Navigation", where we extend the VLN paradigm with a corpus of two-sided, cooperative human-human navigation dialogs! Work with @mmurz, @mayacakmak, and @LukeZettlemoyer.
paper: https://t.co/X69TQSkovJ
code: https://t.co/gyegqAB3yx