Learning to Navigate in Complex Environments

11/11/2016
by   Piotr Mirowski, et al.
0

Learning to navigate in complex environments with dynamic elements is an important milestone in developing AI agents. In this work we formulate the navigation question as a reinforcement learning problem and show that data efficiency and task performance can be dramatically improved by relying on additional auxiliary tasks leveraging multimodal sensory inputs. In particular we consider jointly learning the goal-driven reinforcement learning problem with auxiliary depth prediction and loop closure classification tasks. This approach can learn to navigate from raw sensory input in complicated 3D mazes, approaching human-level performance even under conditions where the goal location changes frequently. We provide detailed analysis of the agent behaviour, its ability to localise, and its network activity dynamics, showing that the agent implicitly learns key navigation abilities.

READ FULL TEXT

page 2

page 15

research
06/02/2021

Robot in a China Shop: Using Reinforcement Learning for Location-Specific Navigation Behaviour

Robots need to be able to work in multiple different environments. Even ...
research
11/15/2022

Agent-State Construction with Auxiliary Inputs

In many, if not every realistic sequential decision-making task, the dec...
research
10/07/2021

How to Sense the World: Leveraging Hierarchy in Multimodal Perception for Robust Reinforcement Learning Agents

This work addresses the problem of sensing the world: how to learn a mul...
research
12/06/2021

Hierarchical Reinforcement Learning with Timed Subgoals

Hierarchical reinforcement learning (HRL) holds great potential for samp...
research
11/28/2019

Playing Games in the Dark: An approach for cross-modality transfer in reinforcement learning

In this work we explore the use of latent representations obtained from ...
research
06/23/2021

Evolving Hierarchical Memory-Prediction Machines in Multi-Task Reinforcement Learning

A fundamental aspect of behaviour is the ability to encode salient featu...
research
03/28/2018

Unsupervised Predictive Memory in a Goal-Directed Agent

Animals execute goal-directed behaviours despite the limited range and s...

Please sign up or login with your details

Forgot password? Click here to reset