The Surprising Effectiveness of Visual Odometry Techniques for Embodied PointGoal Navigation

Abstract

It is fundamental for personal robots to reliably navigate to a specified goal. To study this task, PointGoal navigation has been introduced in simulated Embodied AI environments. Recent advances solve this PointGoal navigation task with near-perfect accuracy (99.6% success) in photo-realistically simulated environments, assuming noiseless egocentric vision, noiseless actuation, and most importantly, perfect localization. However, under realistic noise models for visual sensors and actuation, and without access to a "GPS and Compass sensor," the 99.6%-success agents for PointGoal navigation only succeed with 0.3%. In this work, we demonstrate the surprising effectiveness of visual odometry for the task of PointGoal navigation in this realistic setting, i.e., with realistic noise models for perception and actuation and without access to GPS and Compass sensors. We show that integrating visual odometry techniques into navigation policies improves the state-of-the-art on the popular Habitat PointNav benchmark by a large margin, improving success from 64.5% to 71.7% while executing 6.4 times faster.

ICCV 2021

References in corpus (10)

• Improving neural networks by preventing co-adaptation of feature detectors
• DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent Convolutional Neural Networks
• The Replica Dataset: A Digital Replica of Indoor Spaces
• Matterport3D: Learning from RGB-D Data in Indoor Environments
• MINOS: Multimodal Indoor Simulator for Navigation in Complex Environments
• A Survey on Deep Learning for Localization and Mapping: Towards the Age of Spatial Machine Intelligence
• PyRobot: An Open-source Robotics Framework for Research and Benchmarking
• Cooperative Exploration for Multi-Agent Deep Reinforcement Learning
• Embodied Question Answering in Photorealistic Environments with Point Cloud Perception
• High-Throughput Synchronous Deep RL