KITTI Computer Vision Algorithm Evaluation Dataset
Date
Size
Organization
Publish URL
License
CC BY-NC-SA 3.0
Categories
KITTI is a set of computer vision algorithm evaluation data sets, which are mainly used for related tests in autonomous driving scenarios. The evaluation types include stereo images, optical flow, visual ranging, 3D object detection and 3D tracking. KITTI contains real image data collected in urban, rural and highway scenes. Each image has a maximum of 15 vehicles and 30 pedestrians, and has different degrees of occlusion and truncation.
The dataset consists of 389 pairs of stereo images and optical flow maps, 39.2km of visual ranging sequences, and more than 200k 3D annotated object images, and is sampled and synchronized at 10Hz. The original dataset is divided into five categories: "Road", "City", "Residential", "Campus" and "Person", while 3D object detection is divided into car, van, truck, pedestrian, pedestrian (sitting), cyclist, tram and misc.
The KITTI dataset was jointly released by the Karlsruhe Institute of Technology in Germany and Toyota Research Institute of America in 2013. The related paper is "Vision meet Robotics: The KITTI Dataset".