Agile Robot Navigation through Hallucinated Learning and Sober Deployment

10/16/2020
by   Xuesu Xiao, et al.
0

Learning from Hallucination (LfH) is a recent machine learning paradigm for autonomous navigation, which uses training data collected in completely safe environments and adds numerous imaginary obstacles to make the environment densely constrained, to learn navigation planners that produce feasible navigation even in highly constrained (more dangerous) spaces. However, LfH requires hallucinating the robot perception during deployment to match with the hallucinated training data, which creates a need for sometimes-infeasible prior knowledge and tends to generate very conservative planning. In this work, we propose a new LfH paradigm that does not require runtime hallucination—a feature we call "sober deployment"—and can therefore adapt to more realistic navigation scenarios. This novel Hallucinated Learning and Sober Deployment (HLSD) paradigm is tested in a benchmark testbed of 300 simulated navigation environments with a wide range of difficulty levels, and in the real-world. In most cases, HLSD outperforms both the original LfH method and a classical navigation planner.

READ FULL TEXT
research
07/28/2020

Toward Agile Maneuvers in Highly Constrained Spaces: Learning from Hallucination

While classical approaches to autonomous robot navigation currently enab...
research
08/22/2021

From Agile Ground to Aerial Navigation: Learning from Learned Hallucination

This paper presents a self-supervised Learning from Learned Hallucinatio...
research
08/31/2020

Benchmarking Metric Ground Navigation

Metric ground navigation addresses the problem of autonomously moving a ...
research
04/03/2023

Data-Efficient Policy Selection for Navigation in Partial Maps via Subgoal-Based Abstraction

We present a novel approach for fast and reliable policy selection for n...
research
08/08/2021

Towards real-world navigation with deep differentiable planners

We train embodied neural networks to plan and navigate unseen complex 3D...
research
12/13/2022

Learning Robotic Navigation from Experience: Principles, Methods, and Recent Results

Navigation is one of the most heavily studied problems in robotics, and ...
research
11/01/2020

APPLI: Adaptive Planner Parameter Learning From Interventions

While classical autonomous navigation systems can typically move robots ...

Please sign up or login with your details

Forgot password? Click here to reset