Datasets:

Hilti-Research
/

slam-challenge-2021

text stringlengths 51 146
1628499832.06 1001.250299 998.252862 99.98365 0 0 0 1
1628499850.57 1011.378593 995.372293 99.984625 0 0 0 1
1628499868.09 1017.856449 989.062068 100.00095 0 0 0 1
1628499891.02 1032.931824 986.10322 99.98707 0 0 0 1
1628499931.77 1028.62404 975.886227 99.998824 0 0 0 1
1630485587.76 1013.719242 966.62629 99.742936 0 0 0 1
1630485642.09 1012.240052 970.407837 99.732662 0 0 0 1
1630485653.27 1010.643177 971.494528 99.737397 0 0 0 1
1630485667.37 1011.602591 974.697936 99.735491 0 0 0 1
1630485681.37 1008.541794 980.522521 99.731732 0 0 0 1
1630485693.52 1006.161864 985.082473 99.735494 0 0 0 1
1630485709.09 1005.721868 990.795564 99.730486 0 0 0 1
1630485721.2 1002.804108 995.275297 99.729898 0 0 0 1
1630485735.2 1002.034776 998.304484 99.726576 0 0 0 1
1630485762.43 995.693969 1000.009929 99.724625 0 0 0 1
1630485777.05 994.407726 997.20863 99.721911 0 0 0 1
1630485797.29 999.859247 1002.99055 99.724244 0 0 0 1
1630485810.11 996.676458 1008.743871 99.729808 0 0 0 1
1630485824.94 994.220336 1015.145668 99.729955 0 0 0 1
1630485843.97 991.172393 1025.288325 99.727279 0 0 0 1
1630485859.68 988.045112 1032.506841 99.726017 0 0 0 1
1630485876.37 986.848975 1037.539149 99.730796 0 0 0 1
1630486029.98 984.484212 1042.694624 99.724071 0 0 0 1
1630486112.78 986.220773 1037.372826 99.72629 0 0 0 1
1630486124.61 988.25266 1035.094728 99.725014 0 0 0 1
1630486167.73 992.171399 1025.203501 99.725655 0 0 0 1
1630486181.76 992.678294 1017.781298 99.730951 0 0 0 1
1630486193.57 996.400816 1014.017166 99.729731 0 0 0 1
1630486208.39 997.026379 1006.849925 99.723427 0 0 0 1
1630486220.36 1000.728535 1001.941002 99.725384 0 0 0 1
1630486240.42 996.825288 1000.113588 99.72892 0 0 0 1
1630486254.7 998.549762 998.487503 99.731036 0 0 0 1
1630486272.18 994.126622 997.063118 99.720282 0 0 0 1
1630486290.68 1002.149934 996.293914 99.732853 0 0 0 1
1630486304.51 1005.156242 990.464247 99.732206 0 0 0 1
1630486318.57 1007.613094 983.291861 99.738023 0 0 0 1
1630486333.48 1011.043834 977.45189 99.735978 0 0 0 1
1630486347.75 1010.391088 971.339941 99.740054 0 0 0 1
1630486361.91 1014.311428 966.88315 99.739812 0 0 0 1
1629803884.09 1005.961519 1006.346835 99.912988 0 0 0 1
1629803906.22 1014.36627 1013.349261 100.093062 0 0 0 1
1629803936.55 1011.091547 1007.171031 99.987649 0 0 0 1
1629803958.48 1017.848926 998.395457 100.108362 0 0 0 1
1629803983.9 1034.746536 997.884556 100.480521 0 0 0 1
1629804010.38 1033.506908 982.974289 100.507098 0 0 0 1
1629804032.97 1018.169955 990.691557 100.048361 0 0 0 1
1629804059.6 1009.438269 978.364683 99.820139 0 0 0 1
1629804081.47 1000.758364 989.449666 99.617915 0 0 0 1
1629804107.64 996.164925 997.830949 99.52562 0 0 0 1
1629804136.39 990.469734 1010.342882 99.490306 0 0 0 1
1629804183.21 979.269226 1025.019807 99.293615 0 0 0 1
1629804564.71 986.659825 1024.547994 99.406983 0 0 0 1
1629804583.38 992.226389 1015.980094 99.489552 0 0 0 1
1629804598.67 997.287422 1009.88738 99.600621 0 0 0 1
1629804614.0 1003.567583 1006.396797 99.713727 0 0 0 1
1629804631.39 1010.338106 999.202025 99.8752 0 0 0 1
1629804649.72 1018.316 999.247713 100.133538 0 0 0 1
1629804678.33 1027.136691 999.327775 100.350497 0 0 0 1
1629804704.99 1038.153553 996.793944 100.526519 0 0 0 1
1629804727.99 1038.168478 996.802712 100.525932 0 0 0 1
1629804759.61 1037.559052 986.888682 100.498358 0 0 0 1
1629804776.0 1031.376543 980.980486 100.509723 0 0 0 1
1629804794.39 1021.695496 987.956976 100.261016 0 0 0 1
1629804806.13 1017.688635 990.697189 100.029814 0 0 0 1
1629804828.49 1015.255821 984.111539 99.882457 0 0 0 1
1629804849.66 1022.07178 974.973754 100.375682 0 0 0 1
1629804868.36 1015.23163 968.462071 100.461354 0 0 0 1
1629804887.21 1009.540969 975.915637 100.477851 0 0 0 1
1629804908.04 1001.020464 988.561118 99.617848 0 0 0 1
1629804924.45 998.722372 994.745971 99.584941 0 0 0 1
1629806360.34 998.989890 1003.081681 99.754976 0 0 0 1
1629806379.35 1006.867188 1009.229693 99.929253 0 0 0 1
1629806399.79 1007.775204 1017.754853 99.632330 0 0 0 1
1629806421.91 1004.677557 1026.368611 99.576422 0 0 0 1
1629806438.93 998.607753 1029.390075 99.499761 0 0 0 1
1629806466.56 986.525813 1028.289044 99.778556 0 0 0 1
1629806481.98 988.278394 1021.462977 99.800103 0 0 0 1
1629806498.6 983.815325 1016.627545 99.662062 0 0 0 1
1629806526.54 979.381706 1008.086357 99.562166 0 0 0 1
1629806547.06 979.160973 995.869762 99.577100 0 0 0 1
1629807119.4 997.48704 997.466462 100.344195 0 0 0 1
1629807145.5 999.554445 1007.470168 99.587896 0 0 0 1
1629807173.62 1008.533453 1010.904206 99.827601 0 0 0 1
1629807191.91 1009.741137 1018.893005 99.613926 0 0 0 1
1629807212.73 1006.84959 1026.745225 99.520017 0 0 0 1
1629807241.26 1001.923007 1028.224233 99.512649 0 0 0 1
1629807292.03 986.710904 1028.022915 99.787632 0 0 0 1
1629807312.0 991.812411 1016.421856 99.784072 0 0 0 1
1629807328.72 991.136993 1009.685316 99.333455 0 0 0 1
1629807340.87 988.608236 1006.020292 99.798884 0 0 0 1
1629807361.7 981.618459 996.532809 99.627693 0 0 0 1
1629807380.7 979.395866 984.007529 99.560858 0 0 0 1
1629807400.55 979.99104 974.301264 99.196951 0 0 0 1
1629807418.88 983.277985 964.453184 98.865173 0 0 0 1
1629807438.08 993.449602 962.698356 99.7158 0 0 0 1
1629807454.63 999.902221 965.088025 100.263953 0 0 0 1
1629807472.9 1000.605162 977.010329 100.255964 0 0 0 1
1629807494.17 1002.077696 992.286527 100.193617 0 0 0 1
1629807508.09 1004.39803 995.063879 100.616612 0 0 0 1
1628499383.827698469 1013.672295 1017.153271 99.984098 0 0 0 1

End of preview. Expand in Data Studio

Hilti SLAM Challenge 2021

The Hilti SLAM Challenge 2021 dataset is a multimodal robotics benchmark for evaluating simultaneous localization and mapping (SLAM) and sensor-fusion systems in realistic, difficult environments.

It contains synchronized recordings from a handheld sensor rig captured in indoor and outdoor locations, including offices, laboratories, basements, parking areas, campuses, and construction sites. The sequences feature practical SLAM challenges such as weak geometric or visual features, illumination changes, multi-level motion, and long trajectories.

For detailed sensor specifications, sequence descriptions, calibration notes, and benchmark information, see the official challenge dataset page. Questions and implementation discussions can be raised in the challenge GitHub repository.

Dataset contents

This Hugging Face repository provides approximately 198 GB of data organized as follows:

.
├── rosbags/          # One ROS bag per recorded sequence
├── ground_truth/     # Sparse 3-DoF or continuous 6-DoF reference trajectories
├── calibration.yaml  # Sensor calibration parameters
└── sensor_rig.stp    # CAD model of the handheld sensor rig

The 12 included sequences cover:

indoor office, laboratory, basement, and tracking-area environments;
outdoor campus and construction-site environments; and
a parking-deck sequence.

The ROS bags contain synchronized camera, LiDAR, and inertial measurements, together with the sensor transform tree. Data are stored using ROS message types, including sensor_msgs/Image, sensor_msgs/Imu, and sensor_msgs/PointCloud2; Livox measurements use livox_ros_driver/CustomMsg.

Ground truth

Depending on the sequence, the repository includes either:

sparse 3-DoF ground truth, measured with a total station while the platform was stationary; or
continuous 6-DoF ground truth, recorded with a motion-capture system.

Ground-truth filenames indicate the reference frame, such as pole, prism, or imu. Users should inspect the calibration and TF information before comparing estimated and reference trajectories.

Usage notes

The repository contains raw ROS bags rather than samples exposed through the Hugging Face datasets table interface. Download only the sequences required for your experiment, as individual files are several gigabytes in size.

The data were recorded for ROS 1 workflows. Some message types, particularly the Livox custom message, may require the corresponding ROS driver or a custom parser. The official dataset page also recommends decompressing bags before playback when higher playback speed is needed.

Citation

When using this dataset in academic work, please cite:

@misc{2109.11316,
  author = {Michael Helmberger and Kristian Morin and Beda Berner and Nitish Kumar and Danwei Wang and Yufeng Yue and Giovanni Cioffi and Davide Scaramuzza},
  title = {The {Hilti} {SLAM} Challenge Dataset},
  year = {2021},
  eprint = {arXiv:2109.11316}
}

License

The dataset is released under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 license (CC BY-NC-SA 3.0).

Use is limited to non-commercial purposes. Attribution is required, and adaptations must be distributed under the same or a compatible license. Consult the full license text and the official challenge page before redistribution or derived-data publication.