Learned Point Cloud Geometry Compression

Abstract

This paper presents a novel end-to-end Learned Point Cloud Geometry Compression (a.k.a., Learned-PCGC) framework, to efficiently compress the point cloud geometry (PCG) using deep neural networks (DNN) based variational autoencoders (VAE). In our approach, PCG is first voxelized, scaled and partitioned into non-overlapped 3D cubes, which is then fed into stacked 3D convolutions for compact latent feature and hyperprior generation. Hyperpriors are used to improve the conditional probability modeling of latent features. A weighted binary cross-entropy (WBCE) loss is applied in training while an adaptive thresholding is used in inference to remove unnecessary voxels and reduce the distortion. Objectively, our method exceeds the geometry-based point cloud compression (G-PCC) algorithm standardized by well-known Moving Picture Experts Group (MPEG) with a significant performance margin, e.g., at least 60% BD-Rate (Bjontegaard Delta Rate) gains, using common test datasets. Subjectively, our method has presented better visual quality with smoother surface reconstruction and appealing details, in comparison to all existing MPEG standard compliant PCC methods. Our method requires about 2.5MB parameters in total, which is a fairly small size for practical implementation, even on embedded platform. Additional ablation studies analyze a variety of aspects (e.g., cube size, kernels, etc) to explore the application potentials of our learned-PCGC.

13 pages

References in corpus (4)

• Generative and Discriminative Voxel Modeling with Convolutional Neural Networks
• Deep AutoEncoder-based Lossy Geometry Compression for Point Clouds
• NeuralSampler: Euclidean Point Cloud Auto-Encoder and Sampler
• Gated Context Model with Embedded Priors for Deep Image Compression

Cited by in corpus (17)

• OctAttention: Octree-Based Large-Scale Contexts Model for Point Cloud Compression
• GRASP-Net: Geometric Residual Analysis and Synthesis for Point Cloud Compression
• HUMAN4D: A Human-Centric Multimodal Dataset for Motions and Immersive Media
• Patch-Based Deep Autoencoder for Point Cloud Geometry Compression
• Enhancing context models for point cloud geometry compression with context feature residuals and multi-loss
• Efficient Visual Recognition with Deep Neural Networks: A Survey on Recent Advances and New Directions
• Geometric Prior Based Deep Human Point Cloud Geometry Compression
• Dense Pixel-to-Pixel Harmonization via Continuous Image Representation
• VoxelContext-Net: An Octree based Framework for Point Cloud Compression
• Sparse Tensor-based Multiscale Representation for Point Cloud Geometry Compression
• The JPEG Pleno Learning-based Point Cloud Coding Standard: Serving Man and Machine
• Deep Implicit Volume Compression
• Multiscale Point Cloud Geometry Compression
• Implicit Neural Compression of Point Clouds
• DeepCompress: Efficient Point Cloud Geometry Compression
• Video-based Point Cloud Compression Artifact Removal
• A deep perceptual metric for 3D point clouds