Towards Realistic Underwater Dataset Generation and Color Restoration

TL;DR

This paper introduces a domain adaptation approach to generate realistic underwater datasets and improve color restoration in underwater images using a CNN trained on this data, addressing the limitations of synthetic datasets.

Contribution

It proposes a multimodal domain adaptation method to create realistic underwater datasets and trains a CNN for improved color restoration performance.

Findings

Enhanced dataset realism through domain adaptation.

Improved color restoration accuracy on real underwater images.

Open-source code and models available for reproducibility.

Abstract

Recovery of true color from underwater images is an ill-posed problem. This is because the wide-band attenuation coefficients for the RGB color channels depend on object range, reflectance, etc. which are difficult to model. Also, there is backscattering due to suspended particles in water. Thus, most existing deep-learning based color restoration methods, which are trained on synthetic underwater datasets, do not perform well on real underwater data. This can be attributed to the fact that synthetic data cannot accurately represent real conditions. To address this issue, we use an image to image translation network to bridge the gap between the synthetic and real domains by translating images from synthetic underwater domain to real underwater domain. Using this multimodal domain adaptation technique, we create a dataset that can capture a diverse array of underwater conditions. We then train a simple but effective CNN based network on our domain adapted dataset to perform color restoration. Code and pre-trained models can be accessed at https://github.com/nehamjain10/TRUDGCR