1. [Publications](/index.php/publications)
2. GAvatar: Animatable 3D Gaussian Avatars with Implicit Mesh Learning
 
 # GAvatar: Animatable 3D Gaussian Avatars with Implicit Mesh Learning

  ![](/sites/default/files/publications/output_0.gif) 

 Gaussian splatting has emerged as a powerful 3D representation that harnesses the advantages of both explicit (mesh) and implicit (NeRF) 3D representations. In this paper, we seek to leverage Gaussian splatting to generate realistic animatable avatars from textual descriptions, addressing the limitations (e.g., flexibility and efficiency) imposed by mesh or NeRF-based representations. However, a naive application of Gaussian splatting cannot generate high-quality animatable avatars and suffers from learning instability; it also cannot capture fine avatar geometries and often leads to degenerate body parts. To tackle these problems, we first propose a primitive-based 3D Gaussian representation where Gaussians are defined inside posedriven primitives to facilitate animation. Second, to stabilize and amortize the learning of millions of Gaussians, we propose to use neural implicit fields to predict the Gaussian attributes (e.g., colors). Finally, to capture fine avatar geometries and extract detailed meshes, we propose a novel SDF-based implicit mesh learning approach for 3D Gaussians that regularizes the underlying geometries and extracts highly detailed textured meshes. Our proposed method, GAvatar, enables the large-scale generation of diverse animatable avatars using only text prompts. GAvatar significantly surpasses existing methods in terms of both appearance and geometry quality, and achieves extremely fast rendering (100 fps) at 1K resolution.



 ## Authors



[Ye Yuan](/index.php/person/ye-yuan)

[Xueting Li](/index.php/person/xueting-li)

Yangyi Huang (Chinese University of Hong Kong)

[Shalini De Mello](/index.php/person/shalini-de-mello)

[Koki Nagano](/index.php/person/koki-nagano)

[Jan Kautz](/index.php/person/jan-kautz)

[Umar Iqbal](/index.php/person/umar-iqbal)

 

 

 ## Publication Date



Monday, June 17, 2024

 

 ## Published in



[IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2024](https://openaccess.thecvf.com/content/CVPR2024/papers/Yuan_GAvatar_Animatable_3D_Gaussian_Avatars_with_Implicit_Mesh_Learning_CVPR_2024_paper.pdf)

 

 ## Research Area



[Artificial Intelligence and Machine Learning ](/index.php/research-area/machine-learning-artificial-intelligence)

[Computer Vision](/index.php/research-area/computer-vision)

[Generative AI](/index.php/research-area/generative-ai)

 

 

 ## External Links



[Project Page](https://nvlabs.github.io/GAvatar/)

[ArXiv](https://arxiv.org/abs/2312.11461)

 

 

 ## Uploaded Files



[Paper](https://d1qx31qr3h6wln.cloudfront.net/publications/Yuan_GAvatar_Animatable_3D_Gaussian_Avatars_with_Implicit_Mesh_Learning_CVPR_2024_paper.pdf "Open file in new window")9.2 MB

 

 

 ## Award



Highlight