… 2D feature space to learn depth representations, leading to a … 2D-3D joint methods, our TPVD
restores dense 2D depth in … , 30, 50] and 2D-3D joint [25, 60] depth completion methods. It …

In this paper we introduce a novel approach that harnesses both 2D and 3D attentions to
enable highly accurate depth completion without requiring iterative spatial propagations. …

… Considering that the latent topology information is lost in 2D space, fusing 2D-3D features …
depth completion method, which integrates 3D geometric representation into the 2D learning …

… a depth completion network utilizing PR representation consisting of a shared encoder and
two decoders, where one classifies the pixel’s depth … -to-end depth completion deep learning …

… that depth completion and mesh reconstruction are closely related problems. Inspired by depth
completion … Since our problem focuses on optimizing the mesh representation M, the loss …

… We adopt an encoder-decoder architecture as a simple while effective multi-modality
fusion strategy to learn the joint representation of RGB and depth images, which is utilized to …

… This article aims to depict the development of learning based depth completion … depth
branch are fused by several 2D-3D fusion blocks that jointly learn 2D and 3D representations…

… complete 3D voxel representation of volumetric occupancy and semantic labels from a depth
… tasks, this paper develops a novel iterative learning framework for closed-loop SSC and SS …

… Cross-modal representation learning [47, 48] plays an important role in multi-… joint 2D-3D
representation learning. BPNet [49] proposes a bidirectional projection module to jointly learn …

… Learning-based depth completion methods mainly … depth-plane representation. In this
section, we introduce the multiple-depth-plane representation, how to convert an input depth to the …