Search papers, labs, and topics across Lattice.
This paper introduces HOST, a novel framework that allows robots to acquire new manipulation skills in seconds from a single human video while preserving previously learned skills. By employing a cascade of self-grounded prediction, HOST estimates task progress and translates it into the robot's actions, achieving an average skill acquisition time of 29 seconds and a 62% success rate. Notably, HOST outperforms traditional methods, requiring 50 times fewer demonstrations and acquiring skills 507 times faster than existing approaches.
Robots can learn new skills from just one video in under 30 seconds while retaining what they've already mastered鈥攖ransforming the landscape of robotic skill acquisition.
The ability to acquire skills rapidly and effortlessly while retaining those already mastered is essential for robots. However, current methods still rely on a cumbersome training-time loop that is costly and slow, while eroding skills already mastered. In this paper, we introduce HOST (Human-to-robot One-Shot Skill AcquisiTion), a framework that enables a robot to acquire skills in seconds from a single human video while retaining previously mastered skills. HOST resolves skill acquisition through a cascade of self-grounded prediction. It first estimates the robot's progress within the demonstrated task, then translates the upcoming progression into the robot's own future observations, and finally derives actions from these predicted observations. This cascade is trained on targets coupled to the video demonstration, obtained by mapping the robot trajectory and the video demonstration onto a shared task progress manifold, then redefining each target to align with the future progression of the video. HOST thereby enables the robot to actively follow the demonstrated procedure and adapt it to the robot's embodiment. HOST acquires novel skills at inference time from a single human video in an average of 29 seconds and achieves a 62% average success rate. It exceeds the zero-shot baseline by 45% while retaining previously mastered skills. HOST even exceeds the baseline fine-tuned on 50 robot demonstrations per task while requiring 50 times fewer demonstrations and acquiring each skill 507 times faster. Additional information about HOST is available on the project website.
