Due to distribution shift, deep learning based methods for image dehazing suffer from performance degradation when applied to real-world hazy images. In this paper, this study considers a dehazing framework based on conditional diffusion models for improved generalization to real haze. First, our work finds that optimizing the training objective of diffusion models, i.e., Gaussian noise vectors, is non-trivial. The spectral bias of deep networks hinders the higher frequency modes in Gaussian vectors from being learned and hence impairs the reconstruction of image details. To tackle this issue, this study designs a network unit, named Frequency Compensation block (FCB), with a bank of filters that jointly emphasize the mid-to-high frequencies of an input signal. Our work demonstrates that diffusion models with FCB achieve significant gains in both perceptual and distortion metrics. Second, to further boost the generalization performance, this study proposed a novel data synthesis pipeline, , to augment haze in terms of and . Within the framework, a solid baseline for blind dehazing is set up where models are trained on synthetic hazy-clean pairs, and directly generalize to real data. Extensive evaluations on real dehazing datasets demonstrate the superior performance of the proposed dehazing diffusion model in distortion metrics. Compared to recent methods pre-trained on large-scale, high-quality image datasets, our model achieves a significant PSNR improvement of over 1 dB on challenging databases such as Dense-Haze and Nh-Haze.

中文翻译：

---

实景去雾的频率补偿扩散模型

由于分布变化，基于深度学习的图像去雾方法在应用于现实世界的模糊图像时会出现性能下降的问题。在本文中，本研究考虑了基于条件扩散模型的去雾框架，以改进对真实雾霾的泛化。首先，我们的工作发现优化扩散模型的训练目标（即高斯噪声向量）并非易事。深度网络的谱偏差阻碍了高斯向量中高频模式的学习，从而损害了图像细节的重建。为了解决这个问题，本研究设计了一个名为频率补偿块（FCB）的网络单元，其中包含一组滤波器，共同强调输入信号的中高频。我们的工作表明，使用 FCB 的扩散模型在感知和失真指标方面都取得了显着的进步。其次，为了进一步提高泛化性能，本研究提出了一种新颖的数据合成管道 ，以增强 和 方面的雾霾。在该框架内，建立了盲去雾的坚实基线，其中模型在合成的模糊-干净对上进行训练，并直接推广到真实数据。对真实去雾数据集的广泛评估证明了所提出的去雾扩散模型在畸变指标方面的优越性能。与最近在大规模、高质量图像数据集上预训练的方法相比，我们的模型在 Dense-Haze 和 Nh-Haze 等具有挑战性的数据库上实现了超过 1 dB 的 PSNR 显着改进。