LOBSTgER-enhance: an underwater image enhancement pipeline

TL;DR

This paper introduces LOBSTgER-enhance, a diffusion-based pipeline that effectively restores underwater images by reversing degradations, achieving high perceptual quality and generalization on a small dataset.

Contribution

It presents a novel underwater image enhancement method using synthetic corruption and diffusion models, improving image quality without extensive data.

Findings

High perceptual consistency in restored images

Strong generalization to unseen underwater images

Effective enhancement with a compact model (~11M parameters)

Abstract

Underwater photography presents significant inherent challenges including reduced contrast, spatial blur, and wavelength-dependent color distortions. These effects can obscure the vibrancy of marine life and awareness photographers in particular are often challenged with heavy post-processing pipelines to correct for these distortions. We develop an image-to-image pipeline that learns to reverse underwater degradations by introducing a synthetic corruption pipeline and learning to reverse its effects with diffusion-based generation. Training and evaluation are performed on a small high-quality dataset of awareness photography images by Keith Ellenbogen. The proposed methodology achieves high perceptual consistency and strong generalization in synthesizing 512x768 images using a model of ~11M parameters after training from scratch on ~2.5k images.