Single-Step Latent Diffusion for Underwater Image Restoration

TL;DR

This paper introduces SLURPP, a fast and effective underwater image restoration method that combines latent diffusion models with scene decomposition, significantly improving quality and speed over previous approaches.

Contribution

The paper presents a novel network architecture and synthetic data pipeline for underwater image restoration, enabling state-of-the-art results with much faster processing.

Findings

SLURPP achieves over 200X speedup compared to existing diffusion methods.

It improves PSNR by approximately 3 dB on synthetic benchmarks.

Demonstrates superior qualitative results on real-world underwater images.

Abstract

Underwater image restoration algorithms seek to restore the color, contrast, and appearance of a scene that is imaged underwater. They are a critical tool in applications ranging from marine ecology and aquaculture to underwater construction and archaeology. While existing pixel-domain diffusion-based image restoration approaches are effective at restoring simple scenes with limited depth variation, they are computationally intensive and often generate unrealistic artifacts when applied to scenes with complex geometry and significant depth variation. In this work we overcome these limitations by combining a novel network architecture (SLURPP) with an accurate synthetic data generation pipeline. SLURPP combines pretrained latent diffusion models -- which encode strong priors on the geometry and depth of scenes -- with an explicit scene decomposition -- which allows one to model and account for the effects of light attenuation and backscattering. To train SLURPP we design a physics-based underwater image synthesis pipeline that applies varied and realistic underwater degradation effects to existing terrestrial image datasets. This approach enables the generation of diverse training data with dense medium/degradation annotations. We evaluate our method extensively on both synthetic and real-world benchmarks and demonstrate state-of-the-art performance. Notably, SLURPP is over 200X faster than existing diffusion-based methods while offering ~ 3 dB improvement in PSNR on synthetic benchmarks. It also offers compelling qualitative improvements on real-world data. Project website https://tianfwang.github.io/slurpp/.