Semantic Neural Radiance Fields for Multi-Date Satellite Data

TL;DR

This paper introduces a satellite-specific Neural Radiance Fields model that creates 3D semantic representations from multi-date satellite images, improving robustness and addressing temporal inconsistencies.

Contribution

It presents a novel semantic NeRF model tailored for satellite data, incorporating semantic information to enhance color prediction and handle noisy labels.

Findings

Model effectively improves semantic and color predictions.

Robustness to noisy labels demonstrated.

Provides a new dataset for satellite multi-view semantic analysis.

Abstract

In this work we propose a satellite specific Neural Radiance Fields (NeRF) model capable to obtain a three-dimensional semantic representation (neural semantic field) of the scene. The model derives the output from a set of multi-date satellite images with corresponding pixel-wise semantic labels. We demonstrate the robustness of our approach and its capability to improve noisy input labels. We enhance the color prediction by utilizing the semantic information to address temporal image inconsistencies caused by non-stationary categories such as vehicles. To facilitate further research in this domain, we present a dataset comprising manually generated labels for popular multi-view satellite images. Our code and dataset are available at https://github.com/wagnva/semantic-nerf-for-satellite-data.