IEEE transactions on visualization and computer graphics
Consistent 3D Human Reconstruction from Monocular Video: Learning Correctable Appearance and Temporal Motion Priors.
Cheng Shang, Liang An, TingTing Li, Jiajun Zhang, Yuxiang Zhang, Jidong Tian, Yebin Liu, Xubo Yang
Published: 202510.1109/TVCG.2025.3626741
Abstract
Recent advancements in rendering dynamic humans using NeRF and 3D Gaussian splatting have made significant progress, leveraging implicit geometry learning and image appearance rendering to create digital humans. However, in monocular video rendering,…
Preview only. Read the full abstract at the source