VVC-Compressed Volumetric Video Dataset with Variable Geometry, Texture, and Dynamic Background Configurations

This dataset provides a comprehensive collection of volumetric video sequences compressed using the Versatile Video Coding (VVC\/H.266) standard. It is designed to support the evaluation of compression efficiency, perceptual quality, and multiview rendering in immersive media applications such as Augmented Reality (AR), Virtual Reality (VR), Extended Reality (XR), and 3D telepresence. The dataset is built upon the publicly available 8i Voxelized Full Bodies dataset and comprises two distinct sets: 350 static volumetric sequences (20s duration) generated using four characters under varied compression strategies, and 175 dynamic sequences (12s duration) featuring background motion and multiview camera renderings (left, front, right). Each set explores bitrate\u2013quality trade-offs across multiple quantization levels for geometry and texture, enabling fine-grained analysis of visual artifacts and degradation patterns. The dataset is a valuable benchmark for developing codecs, streaming architectures, and machine learning-based quality predictors in real-time volumetric media workflows.\ud83d\udd17 Repository: https:\/\/github.com\/FHossain5\/VVC-Video-Compression-Dataset