Search

Video-based point cloud compression (V-PCC) has achieved remarkable compression efficiency, which converts point clouds into videos and leverages video codecs for coding. For lossy compression, the undesirable artifacts of attribute images always degrade the point clouds attribute reconstruction quality. In this paper, we propose an Occupancy-assisted Compression Artifact Removal Network (OCARNet) to remove the distortions of V-PCC decoded attribute images for high-quality point cloud attribute reconstruction. Specifically, the occupancy information is fed into network as a prior knowledge to provide more spatial and structural information and to assist in eliminating the distortions of the texture regions. To aggregate the occupancy information effectively, we design a multi-level feature fusion framework with Channel-Spatial Attention based Residual Blocks (CSARB), where the short and long residual connections are jointly employed to capture the local context and long-range dependency. Besides, we propose a Masked Mean Square Error (MMSE) loss function based on the occupancy information to train our proposed network to focus on estimating the attribute artifacts of the occupied regions. To the best of our knowledge, this is the first learning-based attribute artifact removal method for V-PCC. Experimental results demonstrate that our framework outperforms existing state-of-the-art methods and shows the effectiveness on both objective and subjective quality comparisons.

The non-uniform distribution and extremely sparse nature of the LiDAR point cloud (LPC) bring significant challenges to its high-efficient compression. This paper proposes a novel end-to-end, fully-factorized deep framework that represents the original LiDAR point cloud into an octree structure and hierarchically constructs the octree entropy model in layers. The proposed framework utilizes a hierarchical latent variable as side information to encapsulate the sibling and ancestor dependence, which provides sufficient context information for the modeling of point cloud distribution while enabling the parallel encoding and decoding of octree nodes in the same layer. Besides, we propose a residual coding framework for the compression of the latent variable, which explores the spatial correlation of each layer by progressive downsampling, and model the corresponding residual with a fully-factorized entropy model. Furthermore, we propose soft addition and subtraction for residual coding to improve network flexibility. The comprehensive experiment results on the LiDAR benchmark SemanticKITTI and MPEG-specified dataset Ford demonstrate that our proposed framework achieves state-of-the-art performance among all the previous LPC frameworks. Besides, our end-to-end, fully-factorized framework is proved by experiment to be high-parallelized and time-efficient, which saves more than 99.8% of decoding time compared to previous state-of-the-art methods on LPC compression.