Awesome
THQA-3D
Official Access to ACM MM2024：“Subjective and Objective Quality-of-Experience Assessment for 3D Talking Heads”
🌋 System & Background
<p align="center"> <img alt="thqa" src="https://github.com/zyj-2000/THQA-3D/blob/main/pic/system.png" width="92%">🌋 Overview
<p align="center"> <img alt="thqa" src="https://github.com/zyj-2000/THQA-3D/blob/main/pic/overview.png" width="92%">🌋 Distortions
<p align="center"> <img alt="thqa" src="https://github.com/zyj-2000/THQA-3D/blob/main/pic/distortion.png" width="92%">Database Accessible
You can also get THQA-3D dataset from HuggingFace
Frontal Movies : Baidu Yun Pan(code:thqa)
Meta Data (OBJ+UV):
** WARNING: The Meta Data is large so that you must make sure your disk have enough space!!! **
Reference OBJs + UVs + MTLs (2GB): Baidu Yun Pan(code:thqa)
Distorted OBJs + UVs + MTLs Part I - Stream Level Change (5.5GB): Baidu Yun Pan(code:thqa)
Distorted OBJs + UVs + MTLs Part II - Other (11.6GB): Baidu Yun Pan(code:thqa)