LoLI-Street 低光照图像增强数据集

LoLI-Street 是由成均馆大学 (Sungkyunkwan University) 、澳大利亚国立大学 (The Australian National University) 和韩国科技大学 (Tech University of Korea) 的研究团队共同发布的一个专注于低光照图像增强 (LLIE) 的数据集，相关论文成果为「LoLI-Street: Benchmarking Low-Light Image Enhancement and Beyond」，并已被 ACCV’24 接受。这个数据集由来自发达城市街景的 33k 对低光与良好曝光图像组成，涵盖 19k 个目标类别用于目标检测。 LoLI-Street 数据集还包括 1k 张真实低光测试图像，用于在现实条件下测试 LLIE 模型。它对于许多计算机视觉任务至关重要，包括目标检测、跟踪、分割和场景理解。

LoLI-Street is a low-light image enhancement (LLIE)-focused dataset jointly released by research teams from Sungkyunkwan University, The Australian National University, and Tech University of Korea. Its associated paper titled "LoLI-Street: Benchmarking Low-Light Image Enhancement and Beyond" has been accepted by ACCV’24. This dataset consists of 33k pairs of low-light and well-exposed images captured from street scenes in developed cities, covering 19k object categories for object detection tasks. Additionally, LoLI-Street includes 1k real low-light test images for evaluating LLIE models under real-world conditions. It is critical for a wide range of computer vision tasks including object detection, tracking, segmentation, and scene understanding.