Overview of the proposed AquaDiff framework. The forward diffusion process progressively corrupts the clean reference image, while the reverse diffusion process with cross-attention conditioning and chromatic prior guidance recovers the enhanced underwater image.
Underwater images are severely degraded by wavelength-dependent light absorption and scattering, resulting in color distortion, low contrast, and loss of fine details that hinder vision-based underwater applications. To address these challenges, we propose AquaDiff, a diffusion-based underwater image enhancement framework designed to correct chromatic distortions while preserving structural and perceptual fidelity. AquaDiff integrates a chromatic prior-guided color compensation strategy with a conditional diffusion process, where cross-attention dynamically fuses degraded inputs and noisy latent states at each denoising step. An enhanced denoising backbone with residual dense blocks and multi-resolution attention captures both global color context and local details. Furthermore, a novel cross-domain consistency loss jointly enforces pixel-level accuracy, perceptual similarity, structural integrity, and frequency-domain fidelity. Extensive experiments on multiple challenging underwater benchmarks demonstrate that AquaDiff achieves superior color correction and competitive overall image quality across diverse underwater conditions.
Preprocesses degraded images using 3-channel compensation (3C) in Lab color space to correct color casts before diffusion conditioning, with spatially-varying masks to prevent overcompensation.
Dynamically fuses noisy intermediate states with color-compensated conditioning images at each denoising step, enabling adaptive feature weighting based on noise levels.
U-Net architecture with residual dense blocks, dense skip connections, and multi-resolution attention modules to capture both global color context and local structural details.
Novel hybrid loss jointly enforcing pixel-level accuracy, perceptual similarity, structural integrity (SSIM), and frequency-domain fidelity (FFT magnitude spectrum matching).
Results on U90 dataset (TEST-U90) showing superior color correction and detail recovery across diverse underwater scenes.
Results on U45 dataset demonstrating effective correction of green-yellow and blue-green color casts.
Results on C60 dataset for severely degraded images where reference images are unavailable.
Results on S16 dataset with color calibration charts showing precise color restoration.
Quantitative comparison on U45, S16, and C60 datasets using UIQM and UCIQE metrics. AquaDiff achieves the highest UCIQE scores across all datasets, demonstrating superior color correction.
Radar chart illustrating AquaDiff's balanced performance across multiple quality metrics, with consistent superiority in color correction (UCIQE).
| Model Variant | UIQM ↑ | UCIQE ↑ |
|---|---|---|
| Baseline Diffusion Model | 4.12 | 0.486 |
| + CDCL Only | 4.38 | 0.521 |
| + Enhanced U-Net Only | 4.45 | 0.528 |
| AquaDiff (Full Model) | 4.61 | 0.539 |
Ablation studies demonstrating the contribution of Cross-Domain Consistency Loss (CDCL) and Enhanced U-Net backbone.
Superior color restoration: AquaDiff effectively corrects severe color casts ranging from dominant blue-green tints to reddish-brown distortions while preserving fine textures and intricate details.
Haze reduction: AquaDiff achieves superior haze removal with natural appearance, revealing distant objects and fine structural details that remain obscured in competing methods.
@article{shaahid2025aquadiff,
title={AquaDiff: Diffusion-Based Underwater Image Enhancement with Chromatic Prior Guidance and Cross-Domain Consistency},
author={Shaahid, Afrah and Behzad, Muzammil},
journal={arXiv preprint arXiv:XXXX.XXXXX},
year={2025}
}