A sparse autoencoder densifies visual odometry points and feeds them as a geometric prior to self-supervised monocular depth networks, improving depth accuracy on KITTI.
Orb-slam2: an open-source slam system for monocular, stereo and rgb-d cameras
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
abstract
We present ORB-SLAM2 a complete SLAM system for monocular, stereo and RGB-D cameras, including map reuse, loop closing and relocalization capabilities. The system works in real-time on standard CPUs in a wide variety of environments from small hand-held indoors sequences, to drones flying in industrial environments and cars driving around a city. Our back-end based on bundle adjustment with monocular and stereo observations allows for accurate trajectory estimation with metric scale. Our system includes a lightweight localization mode that leverages visual odometry tracks for unmapped regions and matches to map points that allow for zero-drift localization. The evaluation on 29 popular public sequences shows that our method achieves state-of-the-art accuracy, being in most cases the most accurate SLAM solution. We publish the source code, not only for the benefit of the SLAM community, but with the aim of being an out-of-the-box SLAM solution for researchers in other fields.
representative citing papers
MR-SLAM combines passthrough mixed reality with multi-robot SLAM on ROS 2 to let one operator supervise mapping in situ, reporting 8.83 Hz scans, 17.9 m² coverage, and 94.7% occupancy consistency in simulated sessions.
RGB-D visual odometry method performs spatial motion segmentation via grid-based scene flow clustering and temporal tracking with a dual-mode rigid motion model to estimate camera pose from static scene parts in dynamic environments.
citing papers explorer
-
Enhancing self-supervised monocular depth estimation with traditional visual odometry
A sparse autoencoder densifies visual odometry points and feeds them as a geometric prior to self-supervised monocular depth networks, improving depth accuracy on KITTI.
-
Robust Real-time RGB-D Visual Odometry in Dynamic Environments via Rigid Motion Model
RGB-D visual odometry method performs spatial motion segmentation via grid-based scene flow clustering and temporal tracking with a dual-mode rigid motion model to estimate camera pose from static scene parts in dynamic environments.