MORE | Spring 2024

Robot Learning with Adverbial Corrections

Robots programmed to follow human commands have struggled to accurately interpret human intentions down to the movement level. This research proposes a framework for robot learning using natural language adverbial corrections, aiming to bridge the gap between human intent and robotic execution. By integrating natural language instructions, such as “throw faster,” with demonstrations, robots can iteratively adjust their policies based on prior attempts without compromising performance. This approach attempts to enhance human-robot collaboration by enabling robots to understand and execute tasks with language-informed precision, potentially improving assistive robotics making human-robot interactions safer.