Sample Efficient Social Navigation Using Inverse Reinforcement Learning
This addresses the need for robots to navigate safely and appropriately in social spaces, though it is an incremental improvement over existing methods.
The paper tackles the problem of learning socially-compliant navigation policies for mobile robots from human trajectory observations, using an inverse reinforcement learning algorithm enhanced with a replay buffer to improve sample efficiency. The result shows better performance with decreased training time and sample complexity compared to related methods.
In this paper, we present an algorithm to efficiently learn socially-compliant navigation policies from observations of human trajectories. As mobile robots come to inhabit and traffic social spaces, they must account for social cues and behave in a socially compliant manner. We focus on learning such cues from examples. We describe an inverse reinforcement learning based algorithm which learns from human trajectory observations without knowing their specific actions. We increase the sample-efficiency of our approach over alternative methods by leveraging the notion of a replay buffer (found in many off-policy reinforcement learning methods) to eliminate the additional sample complexity associated with inverse reinforcement learning. We evaluate our method by training agents using publicly available pedestrian motion data sets and compare it to related methods. We show that our approach yields better performance while also decreasing training time and sample complexity.