Self-Supervised Monocular Depth Underwater

TL;DR

This paper introduces a self-supervised monocular depth estimation method tailored for underwater environments, overcoming appearance changes and sensor limitations to achieve state-of-the-art results.

Contribution

It adapts self-supervised learning techniques for underwater depth estimation, addressing environment-specific challenges and improving performance without requiring ground truth data.

Findings

Achieved state-of-the-art underwater depth estimation results

Demonstrated effectiveness of self-supervised training in underwater conditions

Improved depth estimation accuracy compared to previous methods

Abstract

Depth estimation is critical for any robotic system. In the past years estimation of depth from monocular images have shown great improvement, however, in the underwater environment results are still lagging behind due to appearance changes caused by the medium. So far little effort has been invested on overcoming this. Moreover, underwater, there are more limitations for using high resolution depth sensors, this makes generating ground truth for learning methods another enormous obstacle. So far unsupervised methods that tried to solve this have achieved very limited success as they relied on domain transfer from dataset in air. We suggest training using subsequent frames self-supervised by a reprojection loss, as was demonstrated successfully above water. We suggest several additions to the self-supervised framework to cope with the underwater environment and achieve state-of-the-art results on a challenging forward-looking underwater dataset.