Ranjansinghhh/UVH-26

UVH-26是一个大规模、专门针对印度交通场景的交通摄像头图像数据集，由印度科学院的AIM团队发布，用于智能交通系统和车辆检测的研究。数据集包含26,646张高分辨率（1080p）图像，这些图像是从班加罗尔约2,800个安全城市监控摄像头在4周时间内采集的。图像通过一场全国性的众包黑客马拉松活动由565名大学生标注，产生了约180万个边界框，涵盖了14种细粒度的车辆类别，这些类别代表了印度交通的实际情况。数据集包含两种不同的标注集：UVH-26-MV（基于多数投票的最终标签）和UVH-26-ST（基于STAPLE算法的标签），以提高标注的可靠性。数据集的结构包括训练集（80%）和验证集（20%），并按照COCO JSON格式组织。

UVH-26 is a large-scale, India-specific traffic-camera image dataset released by AIM @ IISc for research in intelligent transportation systems and vehicle detection. It contains 26,646 high-resolution (1080p) frames sampled from ≈ 2,800 Bengaluru Safe City CCTV cameras over a 4-week period. Images were annotated through a nationwide crowdsourced hackathon involving 565 college students, producing ≈ 1.8 million bounding boxes across 14 fine-grained vehicle classes representative of Indian traffic conditions. The dataset includes two separate annotation sets: UVH-26-MV (final labels computed via majority voting) and UVH-26-ST (labels generated using the STAPLE algorithm) for higher reliability. The dataset structure includes training (80%) and validation (20%) sets, organized in COCO JSON format.