The TUM VI Benchmark for Evaluating Visual-Inertial Odometry

By David Schubert, Thore Goll, Nikolaus Demmel, Vladyslav C. Usenko, Jörg Stückler, Daniel Cremers. IEEE/RJS International Conference on Intelligent RObots and Systems, 2018

479 citations recorded by the source corpus.

Visual odometry and SLAM methods have a large variety of applications in domains such as augmented reality or robotics. Complementing vision sensors with inertial measurements tremendously improves tracking accuracy and robustness, and thus has spawned large interest in the development of visual-inertial (VI) odometry approaches. In this paper, we propose the TUM VI benchmark, a novel dataset with a diverse set of sequences in different scenes for evaluating VI odometry. It provides camera images with 1024×1024 resolution at 20 Hz, high dynamic range and photometric calibration. An IMU measures accelerations and angular velocities on 3 axes at 200 Hz, while the cameras and IMU sensors are time-synchronized in hardware. For trajectory evaluation, we also provide accurate pose ground truth from a motion capture system at high frequency (120 Hz) at the start and end of the sequences which we accurately aligned with the camera and IMU measurements. The full dataset with raw and calibrated data is publicly available. We also evaluate state-of-the-art VI odometry approaches on our dataset.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Geology

View Paper: https://www.semanticscholar.org/paper/fbbc2978758fe459f05d749a0fbd1ef3a3ba3c6c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Geology

Citations in the corpus

Adaptive VIO: Deep Visual-Inertial Odometry with Online Continual Learning (2024, methodology): “In our method, the states from optimization provide feedback signals to the networks, enabling self-supervised online continual learn ing, which has proved effective in EuRoC [5] and VI-TUM datasets [27].”
Are We Ready for Service Robots? The OpenLORIS-Scene Datasets for Lifelong SLAM (2019, background): “A recent contribution is the TUM VI benchmark [5], where aligned visual and IMU data are provided.”
TartanAir: A Dataset to Push the Limits of Visual SLAM (2020, methodology): “On the other hand, The TUM VI dataset [18] and ScanNet dataset [19] cover a diverse set of sequences in different scenes but are obtained with a constant light”
ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual–Inertial, and Multimap SLAM (2020, methodology): “The TUM-VI dataset [82] consists of 28 sequences in 6 different environments, recorded using a hand-held ﬁsheye stereo-inertial rig.”
BAD SLAM: Bundle Adjusted Direct RGB-D SLAM (2019, methodology): “Method Real data RGB-D Stereo Global shutter Sync’ed cameras IMU Accurate GT Geometry GT Benchmark TUM RGB-D [61] X X (1) X (2) TUM VI [57] X X X X X X TUM Mono [12] X X CoRBS [66] X X X X ICL-NUIM [24] X X X X (3) VIPER Odometry [51] X X X InteriorNet [37] X X X X X X X KITTI Odometry [21] X (4) X ”