Abstract: Depth completion is a pivotal challenge in computer vision, aiming at reconstructing the dense depth map from a sparse one, typically with a paired RGB image. Existing learning-based models ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results