Foggy Cityscapes
收藏资源简介:
Foggy Cityscapes数据集由南加州大学创建,用于评估图像去雾方法的性能,特别是对自动驾驶场景中物体检测和分割的影响。数据集包含在雾天条件下拍摄的城市景观图像,旨在帮助研究人员测试和比较各种去雾技术,包括传统滤波器、现代去雾网络、链式变体和视觉语言模型(VLM)图像编辑方法。数据集的应用领域包括自动驾驶感知系统,旨在解决在雾天条件下图像对比度降低和细节模糊的问题,从而提高自动驾驶系统的安全性。
The Foggy Cityscapes dataset was developed by the University of Southern California to evaluate the performance of image dehazing methods, particularly their impacts on object detection and segmentation in autonomous driving scenarios. It includes urban landscape images captured under foggy conditions, aiming to assist researchers in testing and comparing various dehazing techniques, including traditional filters, modern dehazing networks, chain-based variants, and vision-language model (VLM) image editing approaches. This dataset is designed for autonomous driving perception systems, with the goal of addressing issues such as reduced image contrast and blurred details under foggy conditions, thereby enhancing the safety of autonomous driving systems.




