> ML_DATASET // KITTI-VISION-BENCHMARK-SUITE_v1.0
KITTI Vision Benchmark Suite (Geiger et al. 2012)
Karlsruhe Institute of Technology (KIT) / Toyota Technological Institute (TTIC) · Autonomous Driving & Robotics · 6 hours of driving data with synchronized color cameras, Velodyne LiDAR, and GPS/IMU
Autonomous Driving & RoboticsCC-BY-NC-SA-3.06 hours of driving data with synchronized color cameras, Velodyne LiDAR, and GPS/IMUregistration-required
Dataset Profile & Characteristics
Label Type:3D bounding boxes, optical flow, visual odometry, stereo disparities
Languages:en
License Tier:non-commercial
Modalities:image, video, 3d-pointclouds
Intended Use
- Benchmarking stereo vision, optical flow, 3D LiDAR object detection, and visual SLAM
Prohibited / Discouraged Use
- Commercial autonomous driving without additional weather/regional domain training
Bias, Leakage & Privacy Risk Analysis
Privacy / Sensitive Data Risks:
Public street scenes with pedestrians and vehicles in Karlsruhe, Germany.
Known Bias:
Daytime driving in mid-sized European city; sunny to cloudy weather only.
Known Benchmark Leakage:
Official test labels withheld on evaluation server to evaluate 3D object detection fairly.