Shownotes
Autonomous navigation in diverse real-world environments remains one of the greatest challenges in robotics. It demands both robust low-level motion control to handle diverse terrain and high-level reasoning to interpret complex instructions and environments. This paper presents a system that bridges vision-language reasoning and learned control for legged robot navigation in the wild.
Website: share.transistor.fm

