Abstract
Single underwater image enhancement (UIE) is a challenging ill-posed problem, but its development is hindered by two major issues: (1) The labels in underwater reference datasets are pseudo labels, relying on these pseudo ground truths in supervised learning leads to domain discrepancy. (2) Underwater reference datasets are scarce, making training on such small datasets prone to overfitting and distribution shift. To address these challenges, we propose Trans-UIE, a transfer learning-based UIE model that captures the fundamental paradigms of UIE through pretraining and utilizes a dataset composed of both reference and non-reference datasets for fine-tuning. However, fine-tuning the model using only reconstruction loss may introduce confirmation bias. To mitigate this, our method leverages no-reference image quality assessment (NR-IQA) metrics from above-water scenes to guide the transfer learning process across domains while generating enhanced images with the style of the above-water image domain. Additionally, to reduce the risk of overfitting during the pretraining stage, we introduce Pearson correlation loss. Experimental results on both full-reference and no-reference underwater benchmark datasets demonstrate that Trans-UIE significantly outperforms state-of-the-art methods.
Abstract (translated)
单一的水下图像增强（UIE）是一个具有挑战性的、条件不明确的问题，其发展受到两个主要问题的影响：(1) 水下参考数据集中的标签是伪标签，依赖这些伪真实标签进行监督学习会导致领域差异。(2) 水下参考数据集稀缺，使得在这样的小规模数据集上训练容易导致过拟合和分布偏移。为了解决这些问题，我们提出了Trans-UIE，这是一种基于迁移学习的UIE模型，通过预训练捕捉到UIE的基本范式，并使用包含参考和非参考数据集组成的混合数据集进行微调。然而，仅依靠重构损失对模型进行微调可能会引入确认偏差（即只加强已有的假设）。为了解决这个问题，我们的方法利用了来自水面场景的无参考图像质量评估（NR-IQA）指标来指导跨领域的迁移学习过程，并生成具有水面图像领域风格的增强图像。此外，在预训练阶段为了降低过拟合的风险，我们引入了皮尔逊相关损失。 在包含全参考和无参考水下基准数据集上的实验结果表明，Trans-UIE显著优于最先进的方法。
URL
https://arxiv.org/abs/2503.17937
