ADAM-Dehaze: Adaptive Density-Aware Multi-Stage Dehazing for Improved Object Detection in Foggy Conditions

TL;DR

ADAM-Dehaze is an adaptive multi-stage dehazing framework that improves object detection in foggy conditions by classifying haze density and dynamically routing images through specialized processing branches, enhancing visual clarity and detection accuracy.

Contribution

The paper introduces a novel adaptive, density-aware dehazing system that jointly optimizes image restoration and object detection, with a lightweight haze density estimator and dynamic routing for different fog intensities.

Findings

Improves PSNR by up to 2.1 dB on benchmark datasets.

Reduces false alarms by 30 percent.

Increases object detection mAP by up to 13 points.

Abstract

Adverse weather conditions, particularly fog, pose a significant challenge to autonomous vehicles, surveillance systems, and other safety-critical applications by severely degrading visual information. We introduce ADAM-Dehaze, an adaptive, density-aware dehazing framework that jointly optimizes image restoration and object detection under varying fog intensities. A lightweight Haze Density Estimation Network (HDEN) classifies each input as light, medium, or heavy fog. Based on this score, the system dynamically routes the image through one of three CORUN branches: Light, Medium, or Complex, each tailored to its haze regime. A novel adaptive loss balances physical-model coherence and perceptual fidelity, ensuring both accurate defogging and preservation of fine details. On Cityscapes and the real-world RTTS benchmark, ADAM-Dehaze improves PSNR by up to 2.1 dB, reduces FADE by 30 percent, and increases object detection mAP by up to 13 points, while cutting inference time by 20 percent. These results highlight the importance of intensity-specific processing and seamless integration with downstream vision tasks. Code available at: https://github.com/talha-alam/ADAM-Dehaze.