Three-dimensional object tracking and pose estimation represent a pivotal area of research in computer vision, with applications ranging from augmented reality to autonomous navigation and robotics.

Estimating the pose of hand-held objects is a critical and challenging problem in robotics and computer vision. While leveraging multi-modal RGB and depth data is a promising solution, existing ...