Learning from Sparse Demonstrations

TL;DR

This paper introduces Continuous PDP, a method enabling robots to learn objective functions from sparse keyframe demonstrations, effectively handling time misalignments and generalizing to new motion scenarios.

Contribution

The paper presents Continuous PDP, a novel approach for learning robot objectives from sparse, time-stamped keyframes with efficient gradient-based optimization.

Findings

Successfully applied to simulated robot arm and quadrotor

Handles time misalignment between keyframes and execution

Generalizes to unseen motion conditions

Abstract

This paper develops the method of Continuous Pontryagin Differentiable Programming (Continuous PDP), which enables a robot to learn an objective function from a few sparsely demonstrated keyframes. The keyframes, labeled with some time stamps, are the desired task-space outputs, which a robot is expected to follow sequentially. The time stamps of the keyframes can be different from the time of the robot's actual execution. The method jointly finds an objective function and a time-warping function such that the robot's resulting trajectory sequentially follows the keyframes with minimal discrepancy loss. The Continuous PDP minimizes the discrepancy loss using projected gradient descent, by efficiently solving the gradient of the robot trajectory with respect to the unknown parameters. The method is first evaluated on a simulated robot arm and then applied to a 6-DoF quadrotor to learn an objective function for motion planning in unmodeled environments. The results show the efficiency of the method, its ability to handle time misalignment between keyframes and robot execution, and the generalization of objective learning into unseen motion conditions.