CelebV-HQ Test Set
收藏arXiv2025-09-30 收录
下载链接：
http://fadavatar.github.io
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了从开源数据集中随机选取的100个样本，旨在验证模型在真实人类情境下的表现。此外，该数据集还用于定量性能比较，规模为100个样本，其任务是测试谈话式虚拟头像生成的模型性能。
This dataset comprises 100 randomly selected samples sourced from open-source datasets, designed to validate a model's performance in real-world human scenarios. Furthermore, this dataset is employed for quantitative performance comparison, with its core task being to evaluate the performance of models for conversational virtual avatar generation.
提供机构：
Open-source
