Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels

28 Apr 2020Ilya KostrikovDenis YaratsRob Fergus

We propose a simple data augmentation technique that can be applied to standard model-free reinforcement learning algorithms, enabling robust learning directly from pixels without the need for auxiliary losses or pre-training. The approach leverages input perturbations commonly used in computer vision tasks to regularize the value function... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK LEADERBOARD
Continuous Control DeepMind Cheetah Run (Images) DrQ Return 660 # 1
Continuous Control DeepMind Cup Catch (Images) DrQ Return 963 # 1
Continuous Control DeepMind Walker Walk (Images) DrQ Return 921 # 1

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet