GaussianAvatar: Human avatar Gaussian splatting from monocular videos☆

被引:1
|
作者
Lin, Haian
Zhan, Yinwei [1 ]
机构
[1] Guangdong Univ Technol, Guangzhou Univ Town,West Rd, Guangzhou 510006, Peoples R China
来源
COMPUTERS & GRAPHICS-UK | 2025年 / 126卷
关键词
Neural radiance field; 3D Gaussian; Human reconstruction;
D O I
10.1016/j.cag.2024.104155
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Many application fields including virtual reality and movie production demand reconstructing high-quality digital human avatars from monocular videos and real-time rendering. However, existing neural radiance field (NeRF)-based methods are costly to train and render. In this paper, we propose GaussianAvatar, a novel framework that extends 3D Gaussian to dynamic human scenes, enabling fast training and real-time rendering. The human 3D Gaussian in canonical space is initialized and transformed to posed space using Linear Blend Skinning (LBS), based on pose parameters, to learn the fine details of the human body at a very small computational cost. We design a pose parameter refinement module and a LBS weight optimization module to increase the accuracy of the pose parameter detection in the real dataset and introduce multi-resolution hash coding to accelerate the training speed. Experimental results demonstrate that our method outperforms existing methods in terms of training time, rendering speed, and reconstruction quality.
引用
收藏
页数:9
相关论文
共 50 条
  • [41] MVSGaussian: Fast Generalizable Gaussian Splatting Reconstruction from Multi-View Stereo
    Liu, Tianqi
    Wang, Guangcong
    Hu, Shoukang
    She, Liao
    Ye, Xinyi
    Zang, Yuhang
    Cao, Zhiguo
    Li, Wei
    Liu, Ziwei
    COMPUTER VISION-ECCV 2024, PT XVIII, 2025, 15076 : 37 - 53
  • [42] Real-time 3D human pose and motion reconstruction from monocular RGB videos
    Yiannakides, Anastasios
    Aristidou, Andreas
    Chrysanthou, Yiorgos
    COMPUTER ANIMATION AND VIRTUAL WORLDS, 2019, 30 (3-4)
  • [43] iMAPPER: Interaction-guided Scene Mapping from Monocular Videos
    Monszpart, Aron
    Guerrero, Paul
    Ceylan, Duygu
    Yumer, Ersin
    Mitra, Niloy J.
    ACM TRANSACTIONS ON GRAPHICS, 2019, 38 (04):
  • [44] Predicting 4D Hand Trajectory from Monocular Videos
    Ye, Yufei
    Feng, Yao
    Taheri, Omid
    Feng, Haiwen
    Tulsiani, Shubham
    Black, Michael J.
    arXiv,
  • [45] Hand Avatar: Free-Pose Hand Animation and Rendering from Monocular Video
    Chen, Xingyu
    Wang, Baoyuan
    Shum, Heung-Yeung
    2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2023, : 8683 - 8693
  • [46] Robust 3D arm tracking from monocular videos
    Guo, F
    Qian, G
    ADVANCES IN INTELLIGENT COMPUTING, PT 2, PROCEEDINGS, 2005, 3645 : 841 - 850
  • [47] Bags of tricks for learning depth and camera motion from monocular videos
    Dong B.
    Sheng L.
    Virtual Reality and Intelligent Hardware, 2019, 1 (05): : 500 - 510
  • [48] Automated Virtual Navigation and Monocular Localization of Indoor Spaces from Videos
    Wu, Qiong
    Li, Ambrose
    PROCEEDINGS 2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS (CVPRW), 2018, : 1605 - 1613
  • [49] VidSfM: Robust and Accurate Structure-From-Motion for Monocular Videos
    Cui, Hainan
    Tu, Diantao
    Tang, Fulin
    Xu, Pengfei
    Liu, Hongmin
    Shen, Shuhan
    IEEE TRANSACTIONS ON IMAGE PROCESSING, 2022, 31 : 2449 - 2462
  • [50] Pedestrian Segmentation From Uncalibrated Monocular Videos Using a Projection Map
    Jo, Younggwan
    Nam, Woonhyun
    Han, Joon Hee
    IEEE SIGNAL PROCESSING LETTERS, 2009, 16 (07) : 604 - 607