U2NeRF: Unsupervised Underwater Image Restoration and Neural Radiance Fields
Keywords: Underwater Image Restoration, Neural Radiance Fields, Neural Fields in Scattering Medium
TL;DR: A Physics-Based Underwater Restoration Method in a multi-view setup using NeRFs
Abstract: Underwater images suffer from colour shifts, low contrast, and haziness due to light absorption, refraction, scattering and restoring these images has warranted much attention. In this work, we present $\textbf{U2NeRF}$, a transformer-based architecture that learns to render and restore novel views conditioned on multi-view geometry simultaneously. We attempt to implicitly bake restoring capabilities onto the NeRF pipeline and disentangle the predicted color into several components and when combined reconstruct the underwater image in a self-supervised manner. In addition, we release an Underwater View Synthesis $\textbf{UVS}$ dataset consisting of 8 real underwater scenes. Our experiments demonstrate that when optimized on a single scene, U2NeRF outperforms several baselines and showcases improved rendering and restoration capabilities.
Supplementary Material: pdf
Submission Number: 218
Loading