Cycle-consistent Generative Adversarial Networks for Neural Style Transfer using data from Chang'E-4

TL;DR

This paper introduces a cycle-consistent GAN framework for neural style transfer tailored to planetary data from the Chang'E-4 mission, enabling advanced image processing for lunar exploration applications.

Contribution

It presents a novel GAN-based style transfer method specifically designed for planetary imagery, leveraging cycle-consistency to improve transfer quality in space science.

Findings

Effective style transfer on lunar surface images

Enhanced image realism for planetary data

Potential applications in lunar exploration imaging

Abstract

Generative Adversarial Networks (GANs) have had tremendous applications in Computer Vision. Yet, in the context of space science and planetary exploration the door is open for major advances. We introduce tools to handle planetary data from the mission Chang'E-4 and present a framework for Neural Style Transfer using Cycle-consistency from rendered images. The experiments are conducted in the context of the Iris Lunar Rover, a nano-rover that will be deployed in lunar terrain in 2021 as the flagship of Carnegie Mellon, being the first unmanned rover of America to be on the Moon.