Open-source V-PCC project
uvgVPCCenc — practical open-source volumetric video encoder

uvgVPCCenc

uvgVPCCenc is an academic, open-source volumetric video encoder designed as a practical platform for MPEG Video-based Point Cloud Compression (V-PCC)-style coding. It is intended both as a research vehicle for experimenting with new coding tools and as an encoder that can be used by the community for reproducible benchmarking. The project is implemented in C++ and distributed under the permissive BSD-3-Clause license.

Open GitHub repository Project page (UVG) UVG-VPC dataset C++ • BSD-3-Clause • Linux

Features

Practical encoder pipeline

uvgVPCCenc implements a functional end-to-end encoding pipeline inspired by the MPEG V-PCC reference software. It focuses on practical compression and speed, providing a usable baseline for research and deployment.

Volumetric videoPractical pipeline

Input and platform

The encoder accepts point cloud frames as input (geometry as positive integers), supports all voxel sizes, and targets Linux environments. It is written in modern C++ with an open development process.

Point cloud framesVoxel sizesLinux

Research-friendly design

The project is explicitly positioned as a research platform for new coding tools and algorithmic exploration, while remaining straightforward to build and run for reproducible experiments.

Research platformReproducibility

Scope and simplifications

uvgVPCCenc prioritizes practicality by focusing on essential tools; for example, it does not aim to replicate every feature of the full MPEG reference stack (e.g., some reconstruction and lossless-oriented tools are intentionally omitted).

Essential toolsFast iteration

Relation to MPEG V-PCC

Built on the V-PCC ecosystem

MPEG V-PCC (ISO/IEC 23090-5, within the V3C framework) is the standardized solution for efficient compression of dense, surface-like point clouds. uvgVPCCenc is designed to operate in this ecosystem, enabling experimentation and practical encoding aligned with V-PCC concepts.

MPEG V-PCCDense PCCV3C

Complementary to MPEG reference software

MPEG maintains official reference implementations (e.g., TMC2). uvgVPCCenc complements these by providing an independent academic encoder emphasizing practical runtime and ease of experimentation, which can be useful for research prototypes and comparative studies.

Academic implementationTool research

Dataset: UVG-VPC

Point cloud test sequences for benchmarking

The UVG-VPC dataset provides voxelized point cloud sequences intended for development, evaluation, and benchmarking of volumetric video coding pipelines. It is commonly used to report reproducible performance results.

Dataset page: ultravideo.fi/UVG-VPC/

DatasetBenchmarking

Recommended usage

Use UVG-VPC sequences to compare coding tools, encoder configurations, and bitrate/quality tradeoffs. Follow the dataset license terms and citation guidance provided on the dataset page.

Reproducible researchLicense-aware

Paper & citation

Primary publication

Please cite the uvgVPCCenc paper when using the software:

L. Fréneau, G. Gautier, A. Mercat, and J. Vanne, “uvgVPCCenc: Practical open-source encoder for fast V-PCC compression,” in Proc. ACM Multimedia Syst. Conf., Stellenbosch, South Africa, Mar.—Apr. 2025.

Paper (ACM DL): doi:10.1145/3712676.3718336
PDF: tuni.fi PDF mirror

PublicationCitation

This page summarizes a third-party open-source project that uses MPEG V-PCC concepts for research and practical encoding. MPEG standards (ISO/IEC) remain the normative reference; the listed project pages provide licensing and usage terms.