ELITE: Efficient Gaussian Head Avatar from a Monocular Video via Learned Initialization and TEst-time Generative Adaptation

TL;DR

ELITE is a novel system that synthesizes high-fidelity, animatable head avatars from monocular videos efficiently by combining learned initialization with test-time generative adaptation, outperforming prior methods in quality and speed.

Contribution

We propose a new Gaussian avatar synthesis framework that integrates a fast initialization model with a test-time diffusion-based refinement, enabling real-time, high-quality avatar generation from monocular videos.

Findings

Achieves 60x faster synthesis than previous 2D generative methods.

Produces superior visual quality and handles challenging expressions.

Demonstrates strong in-the-wild generalization capabilities.

Abstract

We introduce ELITE, an Efficient Gaussian head avatar synthesis from a monocular video via Learned Initialization and TEst-time generative adaptation. Prior works rely either on a 3D data prior or a 2D generative prior to compensate for missing visual cues in monocular videos. However, 3D data prior methods often struggle to generalize in-the-wild, while 2D generative prior methods are computationally heavy and prone to identity hallucination. We identify a complementary synergy between these two priors and design an efficient system that achieves high-fidelity animatable avatar synthesis with strong in-the-wild generalization. Specifically, we introduce a feed-forward Mesh2Gaussian Prior Model (MGPM) that enables fast initialization of a Gaussian avatar. To further bridge the domain gap at test time, we design a test-time generative adaptation stage, leveraging both real and synthetic images as supervision. Unlike previous full diffusion denoising strategies that are slow and hallucination-prone, we propose a rendering-guided single-step diffusion enhancer that restores missing visual details, grounded on Gaussian avatar renderings. Our experiments demonstrate that ELITE produces visually superior avatars to prior works, even for challenging expressions, while achieving 60x faster synthesis than the 2D generative prior method.