SENIOR: Efficient Query Selection and Preference-Guided Exploration in Preference-based Reinforcement Learning

TL;DR

SENIOR introduces an efficient query selection and exploration method in preference-based reinforcement learning, significantly improving human feedback efficiency and policy learning speed in complex robot tasks.

Contribution

The paper proposes a novel Motion-Distinction-based Selection scheme and preference-guided exploration method to enhance feedback efficiency and accelerate policy learning in PbRL.

Findings

Outperforms five existing methods in feedback efficiency and convergence speed.

Effective in both simulated and real-world robot manipulation tasks.

Videos demonstrating results are available online.

Abstract

Preference-based Reinforcement Learning (PbRL) methods provide a solution to avoid reward engineering by learning reward models based on human preferences. However, poor feedback- and sample- efficiency still remain the problems that hinder the application of PbRL. In this paper, we present a novel efficient query selection and preference-guided exploration method, called SENIOR, which could select the meaningful and easy-to-comparison behavior segment pairs to improve human feedback-efficiency and accelerate policy learning with the designed preference-guided intrinsic rewards. Our key idea is twofold: (1) We designed a Motion-Distinction-based Selection scheme (MDS). It selects segment pairs with apparent motion and different directions through kernel density estimation of states, which is more task-related and easy for human preference labeling; (2) We proposed a novel preference-guided exploration method (PGE). It encourages the exploration towards the states with high preference and low visits and continuously guides the agent achieving the valuable samples. The synergy between the two mechanisms could significantly accelerate the progress of reward and policy learning. Our experiments show that SENIOR outperforms other five existing methods in both human feedback-efficiency and policy convergence speed on six complex robot manipulation tasks from simulation and four real-worlds. Videos can be found on our project website: https://2025senior.github.io/