Multimodal Object Detection dataset
收藏资源简介:
VEDAI: The VEDAI dataset comprises 1246 high-resolution RGB and infrared images, containing 3640 objects categorized into 8 common vehicle classes. Each image, with a resolution of 1024 × 1024 pixels, spans diverse terrains and environments. Notably, vehicles occupy only a small portion of the image pixels, making small object detection particularly challenging.DroneVehicle: The DroneVehicle dataset is a large-scale RGB-IR (visible-light and infrared) multimodal dataset designed for vehicle detection in drone imagery. It encompasses a variety of shooting conditions and scenarios, covering different perspectives, altitudes, and lighting conditions. The dataset includes five vehicle categories: car, truck, bus, van, and freight-car.
VEDAI数据集:该数据集包含1246张高分辨率RGB与红外图像,共计3640个被划分为8类常见车辆的目标。单张图像分辨率为1024×1024像素,拍摄场景涵盖多样地形与环境。值得注意的是,车辆仅占图像像素的极小部分,这使得小目标检测任务极具挑战性。 DroneVehicle数据集:该数据集是一款大规模RGB-IR(可见光与红外)多模态数据集,专为无人机影像中的车辆检测任务设计。其涵盖丰富拍摄条件与场景,包含不同拍摄视角、飞行高度与光照环境。数据集共包含5类车辆目标:轿车(car)、卡车(truck)、巴士(bus)、厢式货车(van)以及货运车厢(freight-car)。




