Learning to Navigate in Cities Without a Map

Abstract

Navigating through unstructured environments is a basic capability of intelligent creatures, and thus is of fundamental interest in the study and development of artificial intelligence. Long-range navigation is a complex cognitive task that relies on developing an internal representation of space, grounded by recognisable landmarks and robust visual processing, that can simultaneously support continuous self-localisation ("I am here") and a representation of the goal ("I am going there"). Building upon recent research that applies deep reinforcement learning to maze navigation problems, we present an end-to-end deep reinforcement learning approach that can be applied on a city scale. Recognising that successful navigation relies on integration of general policies with locale-specific knowledge, we propose a dual pathway architecture that allows locale-specific features to be encapsulated, while still enabling transfer to multiple cities. We present an interactive navigation environment that uses Google StreetView for its photographic content and worldwide coverage, and demonstrate that our learning method allows agents to learn to navigate multiple cities and to traverse to target destinations that may be kilometres away. The project webpage http://streetlearn.cc contains a video summarising our research and showing the trained agent in diverse city environments and on the transfer task, the form to request the StreetLearn dataset and links to further resources. The StreetLearn environment code is available at https://github.com/deepmind/streetlearn

arXiv comment17 pages, 16 figures, published at NeurIPS 2018

References in corpus (20)

• IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
• PlaNet - Photo Geolocation with Convolutional Neural Networks
• Learning to Navigate in Complex Environments
• AI2-THOR: An Interactive 3D Environment for Visual AI
• Learning to Execute
• Building Generalizable Agents with a Realistic and Rich 3D Environment
• Emergence of grid-like representations by training recurrent neural networks to perform spatial localization
• Grounded Language Learning in a Simulated 3D World
• Unsupervised Predictive Memory in a Goal-Directed Agent
• Neural SLAM: Learning to Explore with External Memory
• Playing FPS Games with Deep Reinforcement Learning
• Neural Map: Structured Memory for Deep Reinforcement Learning
• Gated-Attention Architectures for Task-Oriented Language Grounding
• Unifying Map and Landmark Based Representations for Visual Navigation
• One-Shot Reinforcement Learning for Robot Navigation with Interactive Replay
• Memory Augmented Control Networks
• Active Neural Localization
• The AdobeIndoorNav Dataset: Towards Deep Reinforcement Learning based Real-world Indoor Robot Visual Navigation
• A Critical Investigation of Deep Reinforcement Learning for Navigation
• Teaching a Machine to Read Maps with Deep Reinforcement Learning

Cited by in corpus (13)

• Contingency-Aware Exploration in Reinforcement Learning
• Guided Feature Transformation (GFT): A Neural Language Grounding Module for Embodied Agents
• REVERIE: Remote Embodied Visual Referring Expression in Real Indoor Environments
• Are You Looking? Grounding to Multiple Modalities in Vision-and-Language Navigation
• Automated Detecting and Placing Road Objects from Street-level Images
• Reinforcement Learning-Based Coverage Path Planning with Implicit Cellular Decomposition
• Navigation Agents for the Visually Impaired: A Sidewalk Simulator and Experiments
• Multimodal Text Style Transfer for Outdoor Vision-and-Language Navigation
• ArraMon: A Joint Navigation-Assembly Instruction Interpretation Task in Dynamic Environments
• Grounding Open-Domain Instructions to Automate Web Support Tasks
• Adaptive Leader-Follower Formation Control and Obstacle Avoidance via Deep Reinforcement Learning
• Learning to Stop: A Simple yet Effective Approach to Urban Vision-Language Navigation
• EnforceNet: Monocular Camera Localization in Large Scale Indoor Sparse LiDAR Point Cloud