PRISM: Rethinking Scattered Atmosphere Reconstruction as a Unified Understanding and Generation Model for Real-world Dehazing

2026-04-08

AI Summary

This work addresses the challenges posed by non-uniform haze distribution, complex illumination from multiple light sources, and the scarcity of paired real-world data in single-image dehazing. To tackle these issues, the authors propose the Proximal Scattering Atmospheric Reconstruction (PSAR) framework, which extends the classical atmospheric scattering model by jointly reconstructing a haze-free image and spatially varying scattering parameters. The method incorporates an online non-uniform haze synthesis strategy and a selective self-distillation mechanism to enable effective domain adaptation using only unpaired real hazy images. By unifying dehazing as a physically guided understanding-and-generation process, PSAR significantly enhances restoration quality and robustness under complex lighting and non-uniform haze conditions, achieving state-of-the-art performance on multiple real-world benchmarks.

Abstract

Real-world image dehazing (RID) aims to remove haze induced degradation from real scenes. This task remains challenging due to non-uniform haze distribution, spatially varying illumination from multiple light sources, and the scarcity of paired real hazy-clean data. In PRISM, we propose Proximal Scattered Atmosphere Reconstruction (PSAR), a physically structured framework that jointly reconstructs the clear scene and scattering variables under the atmospheric scattering model, thereby improving reliability in complex regions and mixed-light conditions. To bridge the synthetic-to-real gap, we design an online non-uniform haze synthesis pipeline and a Selective Self-distillation Adaptation scheme for unpaired real-world scenarios, which enables the model to selectively learn from high-quality perceptual targets while leveraging its intrinsic scattering understanding to audit residual haze and guide self-refinement. Extensive experiments on real-world benchmarks demonstrate that PRISM achieves state-of-the-art performance on RID tasks.

Problem

• real-world image dehazing
• non-uniform haze
• spatially varying illumination
• paired data scarcity

Innovation

• Proximal Scattered Atmosphere Reconstruction
• Unified Understanding and Generation
• Non-uniform Haze Synthesis
• Selective Self-distillation Adaptation
• Real-world Image Dehazing