AIGC-T23DAQA
收藏资源简介:
AIGC-T23DAQA数据库是目前最大的文本到3D资产质量评估数据库,由上海交通大学构建。该数据库包含了通过6种流行的文本到3D资产生成模型,使用170个文本提示生成的969个经过验证的3D资产,以及分别从质量、真实性和文本资产对应性三个方面对这些资产的主观质量评分。该数据库旨在为文本引导的3D生成质量评估问题提供研究基础,并可用于评估生成3D资产的质量,优化模型训练,选择最佳3D资产等应用。
The AIGC-T23DAQA database is currently the largest text-to-3D asset quality assessment database, constructed by Shanghai Jiao Tong University. This database contains 969 validated 3D assets generated from 170 text prompts via 6 popular text-to-3D asset generation models, along with subjective quality ratings for these assets across three dimensions: quality, authenticity, and text-asset correspondence. This database aims to provide a research foundation for text-guided 3D generation quality assessment tasks, and can be applied to scenarios such as evaluating the quality of generated 3D assets, optimizing model training, and selecting optimal 3D assets.




