SLAM Datasets

This file contains information about publicly available datasets suited for monocular, stereo, RGB-D and lidar SLAM.

Last update: 2021/02/04

New College Dataset

• Year: 2009
• Publication: The New College Vision and Laser Data Set
• Available sensors: GPS, odometry, stereo cameras, omnidirectional camera, lidar
• Ground truth: No
• Link (not sure if it is still available)

Rawseeds Datasets

• Year: 2009
• Publication: RAWSEEDS ground truth collection systems for indoor self-localization and mapping
• Available sensors:
• Ground truth: Yes
• Link

Ford Campus Dataset

• Year: 2011
• Publication: Ford campus vision and lidar data set
• Available sensors: GPS/IMU, lidar, omnidirectional camera
• Ground truth: no
• Link

TUM RGB-D Dataset and Benchmark

• Year: 2012
• Publication: A Benchmark for the Evaluation of RGB-D SLAM Systems
• Available sensors: Kinect/Xtion pro RGB-D cameras
• Ground truth: Yes
• Link

KITTI SLAM Datasets

• Year: 2012
• Publication: Are we ready for Autonomous Driving? The KITTI Vision Benchmark Suite
• Available sensors: GPS/IMU, lidar, 2x stereo pairs
• Ground truth: Yes Link

Málaga Stereo and Laser Datasets

• Year: 2013
• Publication: The Málaga urban dataset: High-rate stereo and LiDAR in a realistic urban scenario
• Available sensors: GPS/IMU, lidar, stereo camera
• Ground truth: Yes Link

NCLT Dataset

• Year: 2015
• Publication: University of Michigan North Campus Long-Term Vision and Lidar Dataset
• Available sensors: IMU/GPS, odometry, lidar, omnidirectional camera
• Ground truth: Yes
• Link

KITTI 360

• Year: 2016
• Publication: Semantic Instance Annotation of Street Scenes by 3D to 2D Label Transfer
• Available sensors: IMU/GPS, stereo, side looking cameras, lidar
• Ground truth: Yes
• Link

EuRoc MAV Dataset

• Year: 2016
• Publication: The EuRoC micro aerial vehicle datasets
• Available sensors: IMU, stereo
• Ground truth: Yes, also for point clouds
• Link

Omnidirectional Stereo Dataset

• Year: 2019
• Publication: SweepNet: Wide-baseline Omnidirectional Depth Estimation
• Available sensors: Synthetic omnidirectional camera
• Ground truth: Yes
• Link

ETH 3D

• Year: 2019
• Publication: BAD SLAM: Bundle Adjusted Direct RGB-D SLAM
• Available sensors: IMU, custom RGB-D and stereo cameras
• Ground truth: Yes
• Link

KAIST Urban Dataset

• Year: 2019
• Publication: Complex urban dataset with multi-level sensors from highly diverse urban environments
• Available sensors: GPS/IMU, odometry, stereo cameras, lidar
• Ground truth: No (only results from some SLAM algorithms)
• Link

Newer College Dataset

• Year: 2020
• Publication: The Newer College Dataset: Handheld LiDAR, Inertial and Vision with Ground Truth
• Available sensors: IMU, stereo cameras, lidar
• Ground truth: Yes
• Link

TartanAIR

• Year: 2020
• Publication: TartanAir: A Dataset to Push the Limits of Visual SLAM
• Available sensors: Synthetic stereo and lidar
• Ground truth: Yes
• Link

Ford AV Dataset

• Year: 2020
• Publication: Ford Multi-AV Seasonal Dataset
• Available sensors: GPS/IMU, lidar, 2x stereo pairs (forward and backword), one dashboard camera, 2 side cameras
• Ground truth: Yes
• Link

UASol

• Year: 2021
• Publication: UASOL: A Large-scale High-resolution Stereo Dataset
• Available sensors: GPS, Zed stereo camera
• Ground truth: Poses given by Zed API
• Link

About

Maintainer: Bruno Silva (bruno.silva AT ect.ufrn.br)
Natalnet Laboratory for Perceptual Robotics
Federal University of Rio Grande do Norte (UFRN)