MaskFusion: Real-Time Recognition, Tracking and Reconstruction of Multiple Moving Objects

Co-Fusion: Real-time Segmentation, Tracking and Fusion of Multiple Objects

MD-SLAM: Multi-cue Direct SLAM. Implements the first photometric LiDAR SLAM pipeline, that works withouth any explicit geometrical assumption. Universal approach, working independently for RGB-D and LiDAR.

Code for CVPR 2018 paper --- Texture Mapping for 3D Reconstruction with RGB-D Sensor

Photometric SLAM and Bundle Adjustment for RGB-D and LiDAR in CUDA.

Lightweight Dense Visual Odometry - a header-only C++ library for real-time dense RGB-D odometry.

The package is common Slam combined with DeepLab-V2-CRF Library.

Code for ICRA 2019 work "MID-Fusion Octree-based Object-Level Multi-Instance Dynamic SLAM"

Visual SLAM learning and training

Implemented a RGBD vSLAM in C++

Code for learning SLAMBOOK

Autonomous Driving Turtlebot using BiSeNet segmentation network on RGBD data

Kinect support for ORB-SLAM2

Modified version of RGBDSLAMv2 which from http://github.com/felixendres/rgbdslam_v2/archive/kinetic.zip

Building a RGBD ptam meant to be used for industrial and educational purposes

DVO-SLAM fork for ROS Kinetic. Used in our BMVC 2017 paper "Efficient Online Surface Correction for Real-time Large-Scale 3D Reconstruction".

Code for CVPR 2018 paper --- Texture Mapping for 3D Reconstruction with RGB-D Sensor