# Efficient 3D Gaussian Head Avatars for Edge Devices

arXiv: https://arxiv.org/abs/2610.09821 · HTML: https://mlsanity.com/paper/2610.09821/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 4 of its 52 references that are in the map.
- First posted on arXiv 7 Oct 2026.

## Builds on (most cited first, 4 of 4)

- [Instance Normalization: The Missing Ingredient for Fast Stylization](https://mlsanity.com/paper/1607.08022/) (2016)
- [GGHead: Fast and Generalizable 3D Gaussian Heads](https://mlsanity.com/paper/2406.09377/) (2024)
- [AGORA: Adversarial Generation Of Real-time Animatable 3D Gaussian Head Avatars](https://mlsanity.com/paper/2512.06438/) (2025)
- [Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view Generation](https://mlsanity.com/paper/2605.25220/) (2026)

## Abstract (arXiv)

Generative 3D Gaussian head avatars provide high-quality, efficient rendering, but synthesising the Gaussian representation remains computationally expensive, limiting deployment on resource-constrained and edge devices. We introduce an efficient generator architecture for unconditional 3D Gaussian head synthesis, based on a parameter-efficient synthesis block and depth-wise separable convolutions while retaining style-based conditioning. Our architecture reduces generator complexity without requiring model compression or quantisation. Compared with the baseline model, our approach reduces FLOPs by 94%, parameter count by 70%, and model size by 81%, while maintaining competitive generation quality. We further demonstrate practical CPU inference and browser-based execution on mobile devices using ONNX Runtime, enabling 3D Gaussian avatar synthesis without dedicated GPU hardware or application-specific software. In addition to conventional image-quality metrics, we evaluate multi-view consistency, training cost, and deployment performance. Code, trained models, and evaluation tools will be released publicly.