1: \begin{abstract}

2: \vspace{-1.0em}

3: 

4: Recently, we have witnessed the explosive growth of various volumetric representations in modeling animatable head avatars.

5: However, due to the diversity of frameworks, there is no practical method to support high-level applications like 3D head avatar editing across different representations.

6: In this paper, we propose a generic avatar editing approach that can be universally applied to various 3DMM-driving volumetric head avatars.

7: To achieve this goal, we design a novel expression-aware modification generative model, which enables

8: lift 2D editing from a single image to a consistent 3D modification field.

9: To ensure the effectiveness of the generative modification process,

10: we develop several techniques, including an expression-dependent modification distillation scheme to draw knowledge from the large-scale head avatar 

11: model and 2D %

12: facial texture editing tools, implicit latent space guidance to enhance model convergence, and a segmentation-based loss reweight strategy for fine-grained texture inversion.

13: Extensive experiments demonstrate that our method delivers high-quality and consistent results across multiple expression and viewpoints.

14: Project page: \href{https://zju3dv.github.io/geneavatar/}{https://zju3dv.github.io/geneavatar/}.

15: 

16: 

17: 

18: 

19: 

20: 

21: 

22: 

23: 

24: 

25: 

26: 

27: 

28: 

29: 

30: \end{abstract}

31: