M3DSS-dataset

M3DSS: A Multi-platform, Multi-sensor, and Multi-scenario Dataset for SLAM System

This work aims to provide a high-quality, multi-platform, multi-sensor, and multi-scenario dataset for SLAM researchers, with the following contributions:
(1) Trajectories collected from challenging environments, ranging from indoor rooms to urban roads;
(2) Data from a diverse set of sensors, with a particular emphasis on event cameras;
(3) Accurate ground truth and high sampling rates for validation;
(4) Extensive evaluation to identify and analyze the limitations of existing SLAM algorithms.

If you find this repositorie is helpful in your research, a simple star or citation of our works should be the best affirmation for us.
This repositorie is licensed under MIT license, international license and is provided for academic purpose. If you are interested in our project for commercial purposes, please contact Dr. Xiaoguang Ma for further communication.

Learn more

An overview of M3DSS sequences

Platform Sequence Duration (s) Lenghth (m) Object Motion Lighting Texture Location Ground truth
Driving Urban_day_01 539 1804 dynamic normal strong normal outdoors GNSS-RTK
Urban_day_02 479 1593 dynamic normal strong normal outdoors GNSS-RTK
Urban_night_01 294 1806 dynamic normal weak normal outdoors GNSS-RTK
Urban_night_02 336 1582 dynamic normal weak normal outdoors GNSS-RTK
Quadruped Robot Elevator 175 \ static normal hdr low indoors FAST-LIO2
Underground 193 98 static normal weak normal indoors FAST-LIO2
Indoor_loop 227 101 static normal normal normal indoors FAST-LIO2
Corridor 205 90 static normal hdr low indoors FAST-LIO2
MCR_light 387 107 static normal normal rich indoors NOKOV
MCR_dark 261 85 static normal weak rich indoors NOKOV
MCR_hdr 330 89 static normal hdr rich indoors NOKOV
Parkway_loop_day 360 122 dynamic normal normal normal outdoors GNSS-RTK
Parkway_loop_night 329 123 dynamic normal weak normal outdoors GNSS-RTK
Forest 309 108 static normal normal rich outdoors FAST-LIO2
Handheld Device Elevator 56 \ static normal hdr low indoors FAST-LIO2
Escalator 126 77 static normal normal low in/outdoors FAST-LIO2
Underground 413 454 static normal weak normal indoors FAST-LIO2
Corridor 103 106 static normal hdr low indoors FAST-LIO2
Indoor_loop 226 280 static normal normal normal indoors FAST-LIO2
MCR_normal_light 134 76 static normal normal rich indoors NOKOV
MCR_normal_dark 136 76 static normal weak rich indoors NOKOV
MCR_normal_hdr 167 112 static normal hdr rich indoors NOKOV
MCR_aggressive_trans_light 123 146 static aggressive normal rich indoors NOKOV
MCR_aggressive_trans_dark 99 87 static aggressive weak rich indoors NOKOV
MCR_aggressive_rot_light 111 101 static aggressive normal rich indoors NOKOV
MCR_aggressive_rot_dark 106 95 static aggressive weak rich indoors NOKOV
MCR_aggressive_6dof_light 99 101 static aggressive normal rich indoors NOKOV
MCR_aggressive_6dof_dark 119 143 static aggressive weak rich indoors NOKOV
MCR_aggressive_6dof_hdr 120 146 static aggressive hdr rich indoors NOKOV
Parkway_loop_day 352 420 dynamic normal normal normal outdoors GNSS-RTK
Parkway_loop_night 368 461 dynamic normal weak normal outdoors GNSS-RTK
Parkway_eight_loop_day 187 203 dynamic normal normal normal outdoors GNSS-RTK
Parkway_eight_loop_night 220 206 dynamic normal weak normal outdoors GNSS-RTK
Forest 151 131 dynamic normal normal rich outdoors FAST-LIO2
UGV Elevator 119 \ static normal hdr low indoors FAST-LIO2
Underground 372 439 static normal weak normal indoors FAST-LIO2
Corridor 107 96 static normal hdr low indoors FAST-LIO2
Indoor_aggressive_circle 50 85 static aggressive normal normal indoors FAST-LIO2
Indoor_loop 305 271 static normal normal normal indoors FAST-LIO2
MCR_normal_loop_light 109 64 static normal normal rich indoors NOKOV
MCR_normal_loop_dark 194 162 static normal weak rich indoors NOKOV
MCR_normal_loop_hdr 204 194 static normal hdr rich indoors NOKOV
MCR_aggressive_loop_light 206 174 static aggressive normal rich indoors NOKOV
MCR_aggressive_loop_dark 187 180 static aggressive weak rich indoors NOKOV
MCR_aggressive_loop_hdr 212 202 static aggressive hdr rich indoors NOKOV
Parkway_loop_day 307 395 dynamic normal normal normal outdoors GNSS-RTK
Parkway_loop_night 352 461 dynamic normal weak normal outdoors GNSS-RTK
Parkway_eight_loop_day 223 254 dynamic normal normal normal outdoors GNSS-RTK
Parkway_eight_loop_night 275 265 dynamic normal weak normal outdoors GNSS-RTK
Street_day 1549 2064 dynamic normal normal normal outdoors GNSS-RTK
Street_night 1631 2439 dynamic normal weak normal outdoors GNSS-RTK
UAV MCR_loop_light 101 105 static aggressive normal rich indoors NOKOV
MCR_loop_dark 153 155 static aggressive weak rich indoors NOKOV
MCR_eight_loop_light 146 112 static aggressive normal rich indoors NOKOV
hdr (high dynamic range): the lighting of scenarios fluctuated between brightness and darkness. MCR: Motion Capture Room with NOKOV.


Acknowledge

Thank you to Feng Huang, Yihan Zhong, Weisong Wen, Li-Ta Hsu from the Department of Aeronautical and Aviation Engineering, the Hong Kong Polytechnical University, Hong Kong SAR, China and Weipeng Guan, Peiyu Chen, Zeren Luo, Xinqi Li, Peng Lu from the Adaptive Robotic Controls Lab (ArcLab), Department of Mechanical Engineering, the University of Hong Kong, Hong Kong SAR, China for their assistance in this project.