Using Conditional Generative Adversarial Networks to Generate Ground-Level Views From Overhead Imagery

TL;DR

This paper introduces a novel conditional GAN framework that synthesizes realistic ground-level views from overhead images, enabling applications like land cover classification by leveraging learned ground appearance features.

Contribution

The paper presents a new cGAN architecture with an encoder-decoder generator that synthesizes ground images from overhead imagery and can be adapted for land cover classification.

Findings

Generated ground images are realistic and representative.

The framework improves land cover classification accuracy.

The cGAN effectively captures ground-level appearance from overhead views.

Abstract

This paper develops a deep-learning framework to synthesize a ground-level view of a location given an overhead image. We propose a novel conditional generative adversarial network (cGAN) in which the trained generator generates realistic looking and representative ground-level images using overhead imagery as auxiliary information. The generator is an encoder-decoder network which allows us to compare low- and high-level features as well as their concatenation for encoding the overhead imagery. We also demonstrate how our framework can be used to perform land cover classification by modifying the trained cGAN to extract features from overhead imagery. This is interesting because, although we are using this modified cGAN as a feature extractor for overhead imagery, it incorporates knowledge of how locations look from the ground.