Visual Memory for Robust Path Following

Abstract

Humans routinely retrace paths in a novel environment both forwards and backwards despite uncertainty in their motion. This paper presents an approach for doing so. Given a demonstration of a path, a first network generates a path abstraction. Equipped with this abstraction, a second network observes the world and decides how to act to retrace the path under noisy actuation and a changing environment. The two networks are optimized end-to-end at training time. We evaluate the method in two realistic simulators, performing path following and homing under actuation noise and environmental changes. Our experiments show that our approach outperforms classical approaches and other learning based baselines.

Neural Information Processing Systems (NeurIPS) 2018. Oral Presentation

Cited by in corpus (16)

• Cross-view Semantic Segmentation for Sensing Surroundings
• ViNG: Learning Open-World Navigation with Visual Goals
• Learned Visual Navigation for Under-Canopy Agricultural Robots
• Talk2Nav: Long-Range Vision-and-Language Navigation with Dual Attention and Spatial Memory
• Semantic Visual Navigation by Watching YouTube Videos
• Learning Navigation Subroutines from Egocentric Videos
• Learning to Navigate Using Mid-Level Visual Priors
• Learning to Visually Navigate in Photorealistic Environments Without any Supervision
• Scaling Local Control to Large-Scale Topological Navigation
• Deep Visual MPC-Policy Learning for Navigation
• Deep Visual Odometry with Adaptive Memory
• Variational Monocular Depth Estimation for Reliability Prediction
• Learning Composable Behavior Embeddings for Long-horizon Visual Navigation
• Building Intelligent Autonomous Navigation Agents
• Learning View and Target Invariant Visual Servoing for Navigation
• Learning Synthetic to Real Transfer for Localization and Navigational Tasks