SingingHead
收藏资源简介:
SingingHead是由上海交通大学创建的大型4D数据集,专注于歌唱头部动画。该数据集包含超过27小时的同步歌唱视频、3D面部运动、歌唱音频和背景音乐,来自76个不同个体和8种音乐类型。数据集的创建过程考虑了数据质量和多样性,确保了高分辨率的视频和准确的3D面部扫描。SingingHead数据集的应用领域包括虚拟头像、扩展现实(XR)和娱乐等,旨在解决歌唱特定数据集的缺乏问题,推动歌唱头部动画技术的发展。
SingingHead is a large-scale 4D dataset dedicated to singing head animation, developed by Shanghai Jiao Tong University. It encompasses over 27 hours of synchronized singing videos, 3D facial motions, singing audio and background music, sourced from 76 unique individuals and spanning 8 music genres. The development of this dataset prioritizes both data quality and diversity, guaranteeing high-resolution videos and accurate 3D facial scans. Application scenarios of the SingingHead dataset cover virtual avatars, extended reality (XR), entertainment and other relevant fields, with the goal of addressing the scarcity of singing-specific datasets and advancing the development of singing head animation technologies.

- 1SingingHead: A Large-scale 4D Dataset for Singing Head Animation上海交通大学 · 2023年



