Reproduce “Rethinking Underwater Image Enhancement: Learning through Text-Guided Degradation Curriculum” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Hongri Tian, Wenxue Li, Tian Ye, Sixiang Chen, Zongxian Su, Lujia Wang, Xianying Wang, Xinhu Zheng and 2 more

The authors introduce a method that creates text-described underwater degradations from clean images and trains an enhancer from easier examples toward harder ones. The system uses controllable synthetic data and transmission guidance. On UIEB’s paired test set, it achieves structural similarity (SSIM) of 0.927, compared with 0.923 for Phaseformer plus TCDA; this benchmark result does not establish robustness across real-world water conditions.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.