A benchmark for the evaluation of RGB-D SLAM systems

By Jürgen Sturm, Nikolas Engelhard, Felix Endres, Wolfram Burgard, Daniel Cremers. 2012

4,280 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Geology, 3D Vision & Pose, Robotics & Control

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

In 3D Vision & Pose

In Robotics & Control

Citations in the corpus

RGB-D datasets using microsoft kinect or similar sensors: a survey (2016, methodology): “1 TUM benchmark dataset TUM Benchmark dataset [85] was founded by University of Technology Munich in July 2011.”
ScanNet: Richly-Annotated 3D Reconstructions of Indoor Scenes (2017, background): “The semantic seg-mentations are more than an order of magnitude larger than any previous RGB-D dataset.”
Simitate: A Hybrid Imitation Learning Benchmark (2019, neutral): “many research topics like autonomous driving [14], RGB-D SLAM systems [15], object tracking [13], [37], [38].”
Multiple Interactions Made Easy (MIME): Large Scale Demonstrations Data for Imitation (2018, background): “Similarly the RGBD SLAM dataset [17] has spurred several papers that have improved the state of the art in SLAM.”
Past, Present, and Future of Simultaneous Localization and Mapping: Toward the Robust-Perception Age (2016, positive): “olor & depth images), Stereo (Bi/trinocular stereo images). Topic References Data IMU GPS Labels LIDAR2D LIDAR3D LOC MAP RGB RGBD Stereo Front-ends and back-ends in small / medium scale scenarios [255] X [276] X [202] X [244] X [115, 207] X [108, 274, 275] X [96] X [26] X [218] X Back-ends in sm”
TartanAir: A Dataset to Push the Limits of Visual SLAM (2020, background): “By providing the large-size and multi-modal data, we enable research of visual SLAM in multiple settings, including monocular SLAM, stereo SLAM, RGB-D SLAM, visual LiDAR SLAM, etc.”
iMAP: Implicit Mapping and Positioning in Real-Time (2021, methodology): “For raw camera recordings, we capture RGB-D videos using a hand-held Microsoft Azure Kinect on a wide variety of environments, as well as test on the TUM RGB-D dataset [30] to evaluate camera tracking.”
Fast and Robust Iterative Closest Point (2020, methodology): “To evaluate their performance on real-world problems, we test the methods on the RGB-D SLAM dataset [65], the ETH laser registration dataset [62], and the 3DMatch dataset [66].”
PL-SVO: Semi-direct Monocular Visual Odometry by combining points and line segments (2016, methodology): “[26] We also evaluate the performance of both SVO and PL-SVO approaches in the TUM Dataset [26], which consist of several sequences recorded with an RGB-D camera in diﬀerent environments, as depicted in Figure 7.”
MobileRGBD, an open benchmark corpus for mobile RGB-D related algorithms (2014, neutral): “Few corpora are available on with a mobile robot equipped with an RGB-D sensor.”
Fusion++: Volumetric Object-Level SLAM (2018, methodology): “Figure 1: Fusion++ with pose graph and discovered inventory on the public fr2 desk sequence [33].”
Continuous 3D Perception Model with Persistent State (2025, positive): “Following MonST3R [98], we evaluate camera pose estimation accuracy on Sintel [10], TUM dynamics [72], and Scan-Net [15] datasets.”
GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting (2023, methodology): “To evaluate the performance of GS-SLAM, we conduct experiments on the Replica [28], and TUM-RGBD [30].”
BAD SLAM: Bundle Adjusted Direct RGB-D SLAM (2019, methodology): “[61]), since this focuses on the SLAM (instead of odometry) performance.”
Visual SLAM algorithms: a survey from 2010 to 2016 (2017, neutral): “TUM RGB-D benchmarking dataset provides RGBD image sequences with 6 DoF camera poses [86].”