DeepNav: Learning to Navigate Large Cities

TL;DR

DeepNav introduces a CNN-based algorithm that learns to navigate large cities efficiently using street-view images, leveraging a large dataset and automated supervision to outperform previous methods.

Contribution

It presents a novel deep learning approach for city navigation using street-view images, with automated supervision and large-scale dataset collection.

Findings

DeepNav outperforms previous hand-crafted feature methods.

Automated annotation process eliminates human input.

Effective navigation across multiple cities and destination types.

Abstract

We present DeepNav, a Convolutional Neural Network (CNN) based algorithm for navigating large cities using locally visible street-view images. The DeepNav agent learns to reach its destination quickly by making the correct navigation decisions at intersections. We collect a large-scale dataset of street-view images organized in a graph where nodes are connected by roads. This dataset contains 10 city graphs and more than 1 million street-view images. We propose 3 supervised learning approaches for the navigation task and show how A* search in the city graph can be used to generate supervision for the learning. Our annotation process is fully automated using publicly available mapping services and requires no human input. We evaluate the proposed DeepNav models on 4 held-out cities for navigating to 5 different types of destinations. Our algorithms outperform previous work that uses hand-crafted features and Support Vector Regression (SVR)[19].