Importance Sampling-Guided Meta-Training for Intelligent Agents in Highly Interactive Environments

TL;DR

This paper presents a novel importance sampling-guided meta-training framework for intelligent agents, enhancing their ability to navigate complex, highly interactive environments like T-intersections and roundabouts by balancing training focus across common and extreme scenarios.

Contribution

It introduces an integrated approach combining guided meta reinforcement learning with importance sampling to iteratively optimize training distributions for better generalization in interactive environments.

Findings

Accelerated training convergence in interactive driving scenarios.

Improved agent performance across diverse interaction levels.

Effective balancing of training focus between common and extreme cases.

Abstract

Training intelligent agents to navigate highly interactive environments presents significant challenges. While guided meta reinforcement learning (RL) approach that first trains a guiding policy to train the ego agent has proven effective in improving generalizability across scenarios with various levels of interaction, the state-of-the-art method tends to be overly sensitive to extreme cases, impairing the agents' performance in the more common scenarios. This study introduces a novel training framework that integrates guided meta RL with importance sampling (IS) to optimize training distributions iteratively for navigating highly interactive driving scenarios, such as T-intersections or roundabouts. Unlike traditional methods that may underrepresent critical interactions or overemphasize extreme cases during training, our approach strategically adjusts the training distribution towards more challenging driving behaviors using IS proposal distributions and applies the importance ratio to de-bias the result. By estimating a naturalistic distribution from real-world datasets and employing a mixture model for iterative training refinements, the framework ensures a balanced focus across common and extreme driving scenarios. Experiments conducted with both synthetic and naturalistic datasets demonstrate both accelerated training and performance improvements under highly interactive driving tasks.