Integrating Algorithmic Planning and Deep Learning for Partially Observable Navigation

TL;DR

This paper introduces Navigation Networks, a differentiable neural network approach that combines algorithmic planning and deep learning for robot navigation in partially observable 3D environments, enabling end-to-end training.

Contribution

It presents NavNets, a novel neural network architecture that integrates state estimation, planning, and acting for robot navigation, bridging model-based and model-free methods.

Findings

Successfully trained NavNets in simulation for complex navigation tasks.

Demonstrated the feasibility of end-to-end learning for partially observable navigation.

Unified planning and perception in a single neural network architecture.

Abstract

We propose to take a novel approach to robot system design where each building block of a larger system is represented as a differentiable program, i.e. a deep neural network. This representation allows for integrating algorithmic planning and deep learning in a principled manner, and thus combine the benefits of model-free and model-based methods. We apply the proposed approach to a challenging partially observable robot navigation task. The robot must navigate to a goal in a previously unseen 3-D environment without knowing its initial location, and instead relying on a 2-D floor map and visual observations from an onboard camera. We introduce the Navigation Networks (NavNets) that encode state estimation, planning and acting in a single, end-to-end trainable recurrent neural network. In preliminary simulation experiments we successfully trained navigation networks to solve the challenging partially observable navigation task.