Project on using event camera to do the moving human detection in low-latency (200hz) and high dynamic range fashion.
My research project on the intersection of vision and robotics.
Using semantic understanding from vision to improve the real-robot navigation in a glasswall environment.
I learned and implemented some fundamental SLAM algorithms based on the Robot Mapping lecture series from Dr. Cyrill Stachniss.
After doing self-learning on the videos and lecture notes, I learned the basics of what is SLAM, how to do (classic paradigms) and why (Derivation).
The topics including:
- Odomotery model and Sensor model
- Kalman filter family: EFK, UKF
- (Rao-Blackwellized) Particle filter
- Feature-based mapping and Occupancy grid based mapping
- Graph-based SLAM (front-end, back-end)