Clothed Human Reconstruction

Clothed human reconstruction aims to create realistic 3D models of people wearing clothes from images or videos, crucial for applications like virtual try-ons and animation. Current research focuses on improving the accuracy and robustness of these models, employing techniques like implicit neural representations (e.g., signed distance functions, Gaussian splatting), parametric models (e.g., SMPL-X) for pose estimation, and innovative strategies to handle complex clothing deformations and occlusions. These advancements are driving progress in fields such as computer graphics, virtual reality, and digital fashion, enabling more realistic and detailed human representations in various applications.