Network Randomization: A Simple Technique for Generalization in Deep Reinforcement Learning

TL;DR

This paper introduces a simple randomization technique using a randomized neural network to enhance the generalization of deep reinforcement learning agents across unseen environments, especially in high-dimensional spaces like images.

Contribution

It proposes a novel input perturbation method with a Monte Carlo inference approach, improving robustness and generalization in deep RL beyond existing regularization techniques.

Findings

Significantly outperforms regularization methods in CoinRun, DeepMind Lab, and robotics tasks.

Enables agents to adapt to new environments by learning invariant features.

Reduces variance in predictions through Monte Carlo approximation.

Abstract

Deep reinforcement learning (RL) agents often fail to generalize to unseen environments (yet semantically similar to trained agents), particularly when they are trained on high-dimensional state spaces, such as images. In this paper, we propose a simple technique to improve a generalization ability of deep RL agents by introducing a randomized (convolutional) neural network that randomly perturbs input observations. It enables trained agents to adapt to new domains by learning robust features invariant across varied and randomized environments. Furthermore, we consider an inference method based on the Monte Carlo approximation to reduce the variance induced by this randomization. We demonstrate the superiority of our method across 2D CoinRun, 3D DeepMind Lab exploration and 3D robotics control tasks: it significantly outperforms various regularization and data augmentation methods for the same purpose.