Collection of original sofware related to our research.
pySLAM-D is a SLAM repository for RGB-D images in python that computes the camera trajectory and yields a 3D reconstruction of the scene. This code is suitable for Reinforcement Learning purposes and utilizes existing C++ libraries for fast and robust visual odometry, loop-closure detection, and pose graph optimization. We have tested this code for visual SLAM in the Habitat-Sim environment.
This repository is available here.
We offer a Matlab implementation of algorithms for:
- Density-based clustering (QuickShift);
- Consistent multi-image matching via density-based clustering (QuickMatch).
The files can be downloaded as a zip archive (quickShiftMatching.zip) or accessed through the release git repository on Bitbucket (https://bitbucket.org/tronroberto/quickshiftmatching).
The code is released under the GPLv3 license.
Set of routines for performing optimization on the essential manifold, using the parametrization as a quotient
manifold of $SO(3) \times SO(3)$, where $SO(3)$ is the manifold of 3-D rotations. These routines are offered as part of the Matlab toolbox ManOpt. The first implementation of the software is provided as a reference. The essential manifold is directly provided as ManOpt package since version 2.0.
- Latest official Manopt release: Manopt: A Matlab toolbox for optimization on manifolds
- Original release: MANOPT 1.06b with essential manifold