Embodied Visual Navigation with Automatic Curriculum Learning in Real Environments

Abstract

We present NavACL, a method of automatic curriculum learning tailored to the navigation task. NavACL is simple to train and efficiently selects relevant tasks using geometric features. In our experiments, deep reinforcement learning agents trained using NavACL significantly outperform state-of-the-art agents trained with uniform sampling -- the current standard. Furthermore, our agents can navigate through unknown cluttered indoor environments to semantically-specified targets using only RGB images. Obstacle-avoiding policies and frozen feature networks support transfer to unseen real-world environments, without any modification or retraining requirements. We evaluate our policies in simulation, and in the real world on a ground robot and a quadrotor drone. Videos of real-world results are available in the supplementary material.

References in corpus (20)

• Playing Atari with Deep Reinforcement Learning
• On Evaluation of Embodied Navigation Agents
• Recurrent World Models Facilitate Policy Evolution
• The Replica Dataset: A Digital Replica of Indoor Spaces
• Experience Replay for Continual Learning
• Learning to Navigate in Complex Environments
• Matterport3D: Learning from RGB-D Data in Indoor Environments
• Learning to Explore using Active Neural SLAM
• Sim2Real Predictivity: Does Evaluation in Simulation Predict Real-World Performance?
• MINOS: Multimodal Indoor Simulator for Navigation in Complex Environments
• Automatic Goal Generation for Reinforcement Learning Agents
• Intrinsic Motivation and Automatic Curricula via Asymmetric Self-Play
• Interactive Gibson Benchmark (iGibson 0.5): A Benchmark for Interactive Navigation in Cluttered Environments
• Unifying Map and Landmark Based Representations for Visual Navigation
• Gibson Env: Real-World Perception for Embodied Agents
• Semi-parametric Topological Memory for Navigation
• Mid-Level Visual Representations Improve Generalization and Sample Efficiency for Learning Visuomotor Policies
• Human Instruction-Following with Deep Reinforcement Learning via Transfer-Learning from Text
• The AdobeIndoorNav Dataset: Towards Deep Reinforcement Learning based Real-world Indoor Robot Visual Navigation
• Learning to Visually Navigate in Photorealistic Environments Without any Supervision

Cited by in corpus (7)

• Autonomous Unmanned Aerial Vehicle Navigation using Reinforcement Learning: A Systematic Review
• Focus on Impact: Indoor Exploration with Intrinsic Motivation
• Variational Automatic Curriculum Learning for Sparse-Reward Cooperative Multi-Agent Problems
• Deep Learning for Embodied Vision Navigation: A Survey
• A Platform-Agnostic Deep Reinforcement Learning Framework for Effective Sim2Real Transfer towards Autonomous Driving
• Towards Deviation-Robust Agent Navigation via Perturbation-Aware Contrastive Learning
• Graph Convolutional Memory using Topological Priors