A Semi-supervised Nighttime Dehazing Baseline with Spatial-Frequency Aware and Realistic Brightness Constraint

TL;DR

This paper introduces a semi-supervised nighttime dehazing method that leverages spatial-frequency domain interaction and brightness constraints to improve realism and effectiveness in real-world hazy scenes.

Contribution

It proposes a novel semi-supervised model with spatial-frequency domain interaction and brightness constraints specifically for nighttime dehazing, addressing domain discrepancy and localized haze issues.

Findings

Outperforms state-of-the-art methods on public benchmarks.

Effectively suppresses haze, glow, and noise in nighttime scenes.

Achieves realistic brightness in dehazed images.

Abstract

Existing research based on deep learning has extensively explored the problem of daytime image dehazing. However, few studies have considered the characteristics of nighttime hazy scenes. There are two distinctions between nighttime and daytime haze. First, there may be multiple active colored light sources with lower illumination intensity in nighttime scenes, which may cause haze, glow and noise with localized, coupled and frequency inconsistent characteristics. Second, due to the domain discrepancy between simulated and real-world data, unrealistic brightness may occur when applying a dehazing model trained on simulated data to real-world data. To address the above two issues, we propose a semi-supervised model for real-world nighttime dehazing. First, the spatial attention and frequency spectrum filtering are implemented as a spatial-frequency domain information interaction module to handle the first issue. Second, a pseudo-label-based retraining strategy and a local window-based brightness loss for semi-supervised training process is designed to suppress haze and glow while achieving realistic brightness. Experiments on public benchmarks validate the effectiveness of the proposed method and its superiority over state-of-the-art methods. The source code and Supplementary Materials are placed in the https://github.com/Xiaofeng-life/SFSNiD.