D-YOLO a robust framework for object detection in adverse weather conditions

TL;DR

D-YOLO is a robust object detection framework designed to perform effectively in adverse weather conditions by integrating image restoration and detection tasks through a double-route network with attention feature fusion.

Contribution

The paper introduces a novel double-route network with an attention feature fusion module and a haze-free feature subnetwork, enhancing detection performance in hazy, snowy, and rainy conditions.

Findings

Outperforms state-of-the-art methods on RTTS and FoggyCityscapes datasets.

Effectively bridges the gap between dehazing and detection tasks.

Improves detection accuracy in adverse weather conditions.

Abstract

Adverse weather conditions including haze, snow and rain lead to decline in image qualities, which often causes a decline in performance for deep-learning based detection networks. Most existing approaches attempts to rectify hazy images before performing object detection, which increases the complexity of the network and may result in the loss in latent information. To better integrate image restoration and object detection tasks, we designed a double-route network with an attention feature fusion module, taking both hazy and dehazed features into consideration. We also proposed a subnetwork to provide haze-free features to the detection network. Specifically, our D-YOLO improves the performance of the detection network by minimizing the distance between the clear feature extraction subnetwork and detection network. Experiments on RTTS and FoggyCityscapes datasets show that D-YOLO demonstrates better performance compared to the state-of-the-art methods. It is a robust detection framework for bridging the gap between low-level dehazing and high-level detection.