Task-Driven Underwater Image Enhancement via Hierarchical Semantic Refinement
IEEE Transactions on Image Processing · 2026 · 35 · 42-56
DOI: 10.1109/tip.2025.3647323 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Image Processing, 2026. Verfasst von Meng Yu und 4 weiteren Autor:innen. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 63 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Program of Shanghai Academic Research Leader, Shanghai Leading Talent Program of Eastern Talent Plan.
Abstract (Originalsprache)
Underwater image enhancement (UIE) is crucial for robust marine exploration, yet existing methods prioritize perceptual quality while overlooking irreversible semantic corruption that impairs downstream tasks. Unlike terrestrial images, underwater semantics exhibit layer-specific degradations: shallow features suffer from color shifts and edge erosion, while deep features face semantic ambiguity. These distortions entangle with semantic content across feature hierarchies, where direct enhancement amplifies interference in downstream tasks. Even if distortions are removed, the damaged semantic structures cannot be fully recovered, making it imperative to further enhance corrupted content. To address these challenges, we propose a task-driven UIE framework that redefines enhancement as machine-interpretable semantic recovery rather than mere distortion removal. First, we introduce a multi-scale underwater distortion-aware generator to perceive distortions across feature levels and provide a prior for distortion removal. Second, leveraging this prior and the absence of clean underwater references, we propose a stable self-supervised disentanglement strategy to explicitly separate distortions from corrupted content through CLIP-based semantic constraints and identity consistency. Finally, to compensate for the irreversible semantic loss, we design a task-aware hierarchical enhancement module that refines shallow details via spatial-frequency fusion and strengthens deep semantics through multi-scale context aggregation, aligning results with machine vision requirements. Extensive experiments on segmentation, detection, and saliency tasks demonstrate the superiority of our method in restoring machine-friendly semantics from degraded underwater images. Our code is available at https://github.com/gemyumeng/HSRUIE.
Förderung
- National Natural Science Foundation of China
- Program of Shanghai Academic Research Leader
- Shanghai Leading Talent Program of Eastern Talent Plan
Zitieren
APA
Yu, M., Shen, L., Yu, Y., Zhang, Y., & Le, R. (2026). Task-Driven Underwater Image Enhancement via Hierarchical Semantic Refinement. IEEE Transactions on Image Processing, 35, 42-56. https://doi.org/10.1109/tip.2025.3647323
Vancouver
Yu M, Shen L, Yu Y, Zhang Y, Le R. Task-Driven Underwater Image Enhancement via Hierarchical Semantic Refinement. IEEE Transactions on Image Processing. 2026;35:42-56. doi:10.1109/tip.2025.3647323
BibTeX
@article{Yu2026,
author = {Yu, Meng and Shen, Liquan and Yu, Yihan and Zhang, Yu and Le, Rui},
title = {Task-Driven Underwater Image Enhancement via Hierarchical Semantic Refinement},
journal = {IEEE Transactions on Image Processing},
year = {2026},
volume = {35},
pages = {42--56},
doi = {10.1109/tip.2025.3647323}
}
EUVIMED