Intro
In recent years, static meshes with texture maps have become one of the most prevalent digital representations of 3D shapes in various applications, such as animation, gaming, medical imaging, and cultural heritage applications.
To facilitate the study of static mesh compression algorithm and objective quality metric, we create a large-scale textured mesh database containing rich contents and multiple types of distortion in this paper, called SJTU-TMQA. 21 reference meshes are selected from different categories, including human figures, inanimate objects, animals, and plants. Eight types of distortion: six single distortion types and two superimposed distortion types are injected into each reference mesh at different distortion levels, leading to 945 distorted meshes. The distorted meshes are rendered into processed video sequences (PVS) with a predefined camera path, and 73 viewers aged 18 to 30 are collected to perform subjective experiments with a lab environment.
