M-PCCD (MPEG Point Cloud Compression Dataset)

评估了新兴的 MPEG 点云编解码器（V-PCC 和 G-PCC 变体），并研究了速率分配的最佳实践 [1]。为此，进行了三个实验。在第一个实验中，对编解码器进行了严格的评估，采用小组专家在精心挑选的一组模型上规定的测试条件，使用主观和客观的质量评估方法。在另外两个实验中，对仅几何和几何加颜色编码的不同速率分配方案进行了主观评估，以便就给定比特率的感知质量方面得出最佳性能方法的结论。在此网页中，我们公开了与每个实验评估的刺激相关的质量分数。出于再现性的目的，使用的内容不属于标准化机构采用的既定点云存储库的一部分，将被重新分发。此外，还提供了脚本以生成本研究中使用的参考模型和渲染相关元数据。

We evaluated emerging MPEG point cloud codecs (variants of V-PCC and G-PCC) and investigated best practices for rate allocation [1]. To this end, three experiments were conducted. In the first experiment, the codecs were rigorously evaluated using both subjective and objective quality assessment methods, under test conditions specified by a panel of experts on a carefully selected set of models. In the other two experiments, subjective evaluations were conducted on different rate allocation schemes for geometry-only and geometry-plus-color coding, to draw conclusions regarding the optimal performance approaches in terms of perceived quality at a given bitrate. In this webpage, we have made publicly available the quality scores associated with the stimuli evaluated in each experiment. For the purpose of reproducibility, the content used, which is not part of the established point cloud repositories adopted by standardization bodies, will be redistributed. In addition, scripts for generating the reference models and rendering-related metadata used in this study are also provided.

搜集汇总

背景与挑战

背景概述

M-PCCD数据集专注于评估MPEG点云编解码器的性能，包含三个实验的质量评估结果和速率分配方案研究，适用于点云压缩领域的研究和开发。