• Publications
  • Influence
LSD-SLAM: Large-Scale Direct Monocular SLAM
TLDR
A novel direct tracking method which operates on \(\mathfrak{sim}(3)\), thereby explicitly detecting scale-drift, and an elegant probabilistic solution to include the effect of noisy depth values into tracking are introduced.
Direct Sparse Odometry
TLDR
The experiments show that the presented approach significantly outperforms state-of-the-art direct and indirect methods in a variety of real-world settings, both in terms of tracking accuracy and robustness.
Semi-dense Visual Odometry for a Monocular Camera
We propose a fundamentally novel approach to real-time visual odometry for a monocular camera. It allows to benefit from the simplicity and accuracy of dense tracking - which does not depend on
The Replica Dataset: A Digital Replica of Indoor Spaces
TLDR
Replica, a dataset of 18 highly photo-realistic 3D indoor scene reconstructions at room and building scale, is introduced to enable machine learning (ML) research that relies on visually, geometrically, and semantically realistic generative models of the world.
Large-scale direct SLAM with stereo cameras
TLDR
A novel Large-Scale Direct SLAM algorithm for stereo cameras (Stereo LSD-SLAM) that runs in real-time at high frame rate on standard CPUs, capable of handling aggressive brightness changes between frames - greatly improving the performance in realistic settings.
Camera-based navigation of a low-cost quadrocopter
TLDR
A novel, closed-form solution to estimate the absolute scale of the generated visual map from inertial and altitude measurements and shows its robustness to temporary loss of visual tracking and significant delays in the communication process.
A Photometrically Calibrated Benchmark For Monocular Visual Odometry
TLDR
A novel, simple approach to non-parametric vignette calibration, which requires minimal set-up and is easy to reproduce and thoroughly evaluate two existing methods (ORB-SLAM and DSO) on the dataset.
Direct visual-inertial odometry with stereo cameras
TLDR
This work proposes a novel direct visual-inertial odometry method for stereo cameras that outperforms not only vision-only or loosely coupled approaches, but also can achieve more accurate results than state-of-the-art keypoint-based methods on different datasets, including rapid motion and significant illumination changes.
Large-scale direct SLAM for omnidirectional cameras
TLDR
A real-time, direct monocular SLAM method for omnidirectional or wide field-of-view fisheye cameras, providing a fast yet accurate approach to incremental stereo directly on distorted images and evaluating the framework on real-world sequences taken with a 185° fISheye lens.
...
...