Abstract: Visual understanding of 3-D objects is essential for robotic manipulation, autonomous navigation, and augmented reality. However, existing methods struggle to perform this task efficiently ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback