FreePath：面向图像去雾的神经平均自由程框架
FreePath: Neural Mean Free Path for Image Dehazing
AI 导读
FreePath 是一种用于单图像去雾的神经平均自由程（mean free path）框架，将去雾表示为特征空间中的路径状态恢复，而非估计大气参数。其 Mean Free Path Residual Block 学习路径与碰撞场以引导雾约束校正，FreePath State Conservation Integration 在深层特征聚合中保留恢复的路径状态。
正文 · 原文
Abstract:Single image dehazing requires recovering scene radiance from visual responses that are weakened and veiled by participating media. Existing methods either invert explicit scattering variables, which can be unstable under complex scenes, or learn direct restoration mappings, which often discard the physical organization of haze formation. We present \emph{FreePath}, a neural mean free path framework for image dehazing. Instead of estimating atmospheric parameters, FreePath represents dehazing as feature space path state restoration, inspired by the expected distance a visual response can propagate before interacting with haze. Within a residual encoder and decoder, the Mean Free Path Residual Block learns path and collision fields to guide haze constrained correction, while FreePath State Conservation Integration preserves restored path states during deep feature aggregation. Experiments on paired CDD-11 haze images and real RTTS scenes show that FreePath consistently improves full reference fidelity, no reference perceptual quality, and visual faithfulness over representative dehazing baselines.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.07398 [cs.CV] |
| (or arXiv:2611.07398v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.07398 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Rongrong Gao [view email]
[v1]
Thu, 5 Nov 2026 03:50:56 UTC (4,615 KB)
来源：arXiv cs.AI · arxiv.org