End-to-End United Video Dehazing and Detection

TL;DR

This paper introduces an end-to-end neural network that jointly performs video dehazing and object detection, leveraging temporal consistency for improved accuracy in hazy videos.

Contribution

It proposes the first end-to-end video dehazing network and integrates it with object detection for enhanced performance.

Findings

Improved detection accuracy in hazy videos.

Temporal fusion strategies enhance dehazing quality.

Joint training yields more stable detection results.

Abstract

The recent development of CNN-based image dehazing has revealed the effectiveness of end-to-end modeling. However, extending the idea to end-to-end video dehazing has not been explored yet. In this paper, we propose an End-to-End Video Dehazing Network (EVD-Net), to exploit the temporal consistency between consecutive video frames. A thorough study has been conducted over a number of structure options, to identify the best temporal fusion strategy. Furthermore, we build an End-to-End United Video Dehazing and Detection Network(EVDD-Net), which concatenates and jointly trains EVD-Net with a video object detection model. The resulting augmented end-to-end pipeline has demonstrated much more stable and accurate detection results in hazy video.