S-Avatar: Diffusion-Guided Gaussian Head Avatars from a Single Image

2026-07-30

AI Summary

This work addresses the challenge of 3D inconsistency in novel-view synthesis when generating head avatars from a single image, a limitation that undermines realism in virtual reality applications. To overcome this, the authors propose a three-stage framework that uniquely integrates diffusion model–guided 3D Gaussian Splatting (3DGS) with the FLAME parametric head model. Starting from a single input image, the method first generates an initial 3DGS representation, then aligns it with FLAME’s semantic structure, and finally constructs a spatially anchored template enabling real-time rendering driven by expressions and head poses. Without requiring multi-view inputs, the approach produces high-fidelity, geometrically consistent, and animatable head avatars. Extensive evaluations on public datasets demonstrate significant improvements over state-of-the-art methods in both novel-view synthesis and expression generation, highlighting its suitability for VR/AR applications.

Abstract

We propose S-Avatar, a novel method for generating photorealistic 3D head avatars from a single image using a diffusion-guided 3D model generation module and strategies for animating 3D Gaussian Splatting (3DGS). While single-image head avatar reconstruction is crucial for lifelike Virtual Reality (VR) applications, existing approaches often struggle to preserve 3D consistency under unseen viewpoints. S-Avatar addresses this limitation through a three-stage pipeline. First, a high-resolution 3DGS is synthesized directly from a single image using a diffusion-based Gaussian splat generation module. Next, the parametric head model FLAME is aligned with the generated 3DGS by optimizing its parameters and spatial transformations. Finally, to adapt the 3DGS to FLAME variations, we construct a binding template that encodes the spatial relationship between the initial splats and FLAME. The dynamic 3D head avatar can then be rendered in real time by deforming the 3DGS with the binding template. By combining diffusion-guided canonical 3DGS generation with FLAME-based control, our method achieves efficient and accurate reconstruction with enhanced 3D consistency. Evaluations on public datasets demonstrate that S-Avatar outperforms state-of-the-art methods in novel-view and expression generation, achieving superior realism and consistency. Consequently, our approach represents a significant advance in accessible avatar creation, applicable to a wide range of VR/AR applications. The project page is available at https://github.com/hailsong/savatar.

Problem

• 3D head avatar
• single-image reconstruction
• 3D consistency
• novel-view synthesis
• virtual reality

Innovation

• Diffusion-guided 3D generation
• 3D Gaussian Splatting
• Single-image avatar reconstruction
• FLAME alignment
• Binding template