Understanding Thermal Superpoint Slam
Exploring Thermal Superpoint Slam reveals several interesting facts. Code available here: https://github.com/specarmi/Thermal_SuperPoint_SLAM.
Key Takeaways about Thermal Superpoint Slam
- Visual odometry based on deep learning-based keypoint detector and descriptor.
- Using this github library: https://github.com/magicleap/SuperPointPretrainedNetwork.
- Sparse Depth Enhanced Direct
- requires source YOUR_PATH/catkin_orb/src/ORB_SLAM2/Examples/ROS/ORB_SLAM2/build/devel/setup.bash Sensor: A65, ...
- dataset link: https://sites.google.com/view/rpmsthereo/
Detailed Analysis of Thermal Superpoint Slam
Abstract: Mixed Reality and Robotics require robust Simultaneous Localization and Mapping ( Superpoint 720p Room Sample - 60fps https://arxiv.org/abs/1712.07629.
Illustration of the
Stay tuned for more updates related to Thermal Superpoint Slam.
