MaskUKF: An Instance Segmentation Aided Unscented Kalman Filter for 6D Object Pose and Velocity Tracking

[pattacini]

@xEnVrE from HSP@IIT just got his work on 6D object pose tracking published on Frontiers yesterday 🎉

The framework is called MaskUKF and combines DL-based object segmentation and depth information with a serial Unscented Kalman Filter to track the pose and the velocity of an object in real-time.

Check out the awesome tracking performance we can achieve with MaskUKF, also when applied to a reaching task with the iCub 🎥

masukf.mp4

• Link to the paper: https://www.frontiersin.org/articles/10.3389/frobt.2021.594583/full
• Link to the video: https://www.youtube.com/watch?v=UZ1CGojdxrA&t=3s

cc @lornat75 @fbottarel @claudiofantacci @giuliavezzani

[pattacini]

Code will be made available soon at https://github.com/robotology/mask-ukf.