We are excited to present our latest research paper proposing a large-scale quality assessment database for digital human heads (DHHs). Our database includes a total of 1,540 DHH models, including 55 high-quality reference stimuli and their corresponding manually degraded versions with 7 types of distortions.
We invited 255 human subjects to participate in the DHH scanning experiment. The reference DHH models, which cover both male and female, young and old subjects, are in the format of textured meshes. We carefully selected 55 high-quality generated DHH models as reference stimuli.
The 7 types of distortions we used to degrade the reference DHHs include surface simplification, position compression, UV compression, texture sub-sampling, texture compression, color noise, and geometry noise. We obtained a total of 1,540 distorted DHHs.
A well-controlled subjective experiment was conducted to collect mean opinion scores (MOS) for the distorted DHHs. We obtained the largest subject-rate quality assessment database for digital human heads (DHHQA).
Copyright (c) 2023 Shanghai Jiao Tong University, Institute of Image Communication and Network Engineering. 🌟🌟🌟
This database contains sensitive information and shall not be distributed, used for commercial purposes or modified without permission. If you wish to use this database, please send an email for request and provide your name and affiliation. By sending an email, you agree to comply with the terms of this license.
The use of this database is subject to the following restrictions:
-
It is for personal or internal use only and shall not be used for any public distribution or commercial purposes.
-
It shall not be used for any illegal or unethical purposes.
-
It shall not be modified, cropped, converted, edited, reverse-engineered or otherwise altered in any form.
-
Copyright information shall be acknowledged for any use of this database, and this copyright statement shall be included in any data, graphics, images, photos, audio and/or video files used in any publication or product.
Violations of this license will be subject to legal sanctions.
Shanghai Jiao Tong University, Institute of Image Communication and Network Engineering reserves the right to modify this license at any time without further notice.
The database can be accessed here.
If you find our paper useful, please cite our work as:
@inproceedings{zhang2023perceptual,
title={Perceptual Quality Assessment for Digital Human Heads},
author={Zhang, Zicheng and Zhou, Yingjie and Sun, Wei and Min, Xiongkuo and Wu, Yuzhe and Zhai, Guangtao},
booktitle={IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
year={2023},
organization={IEEE}
}