Driving-Video Dehazing with Non-Aligned Regularization for Safety Assistance

TL;DR

This paper introduces a novel non-aligned regularization method for real driving-video dehazing, effectively handling unaligned hazy and clear frames using reference matching and advanced attention modules, validated on a new dataset.

Contribution

It proposes a pioneering non-aligned regularization strategy with reference matching and attention modules for improved driving-video dehazing in dynamic scenarios.

Findings

Outperforms state-of-the-art dehazing methods on the GoProHazy dataset.

Effectively handles unaligned hazy and clear video pairs.

Demonstrates robustness in diverse rural and urban environments.

Abstract

Real driving-video dehazing poses a significant challenge due to the inherent difficulty in acquiring precisely aligned hazy/clear video pairs for effective model training, especially in dynamic driving scenarios with unpredictable weather conditions. In this paper, we propose a pioneering approach that addresses this challenge through a nonaligned regularization strategy. Our core concept involves identifying clear frames that closely match hazy frames, serving as references to supervise a video dehazing network. Our approach comprises two key components: reference matching and video dehazing. Firstly, we introduce a non-aligned reference frame matching module, leveraging an adaptive sliding window to match high-quality reference frames from clear videos. Video dehazing incorporates flow-guided cosine attention sampler and deformable cosine attention fusion modules to enhance spatial multiframe alignment and fuse their improved information. To validate our approach, we collect a GoProHazy dataset captured effortlessly with GoPro cameras in diverse rural and urban road environments. Extensive experiments demonstrate the superiority of the proposed method over current state-of-the-art methods in the challenging task of real driving-video dehazing. Project page.