SMPLitex
收藏资源简介:
SMPLitex数据集是由胡安卡洛斯国王大学马德里分校的研究团队创建,专注于从单张图像中估计和操纵完整的3D人体外观。该数据集包含100个精选的高质量3D人体纹理样本,这些样本是通过对SMPLitex模型进行条件采样生成的,涵盖了多种服装类型、身份和面部细节。创建过程中,研究团队利用了最新的扩散模型技术,确保了纹理的多视角一致性和3D感知能力。SMPLitex数据集的应用领域广泛,包括3D姿态估计、人体重建和神经渲染,旨在解决现有数据集在细节、多样性和样本数量上的不足。
The SMPLitex dataset was developed by a research team from the Madrid campus of Universidad Rey Juan Carlos, focusing on estimating and manipulating full 3D human appearance from a single monocular image. It contains 100 carefully selected high-quality 3D human texture samples, which are generated via conditional sampling on the SMPLitex model, encompassing diverse clothing types, individual identities, and fine facial details. During the development process, the research team adopted state-of-the-art diffusion model technologies to ensure multi-view consistency and 3D perceptual quality of the textures. The SMPLitex dataset has broad application scenarios including 3D pose estimation, human reconstruction and neural rendering, and aims to address the limitations of existing datasets in terms of detail richness, diversity and sample quantity.

- 1SMPLitex: A Generative Model and Dataset for 3D Human Texture Estimation from Single Image胡安卡洛斯国王大学马德里分校 · 2023年



