Recent advancements in automatic 3D avatar generation guided by text have made significant progress. However, existing methods have limitations such as oversaturation and low-quality output. To address these challenges, we propose X-Oscar, a progressive framework for generating high-quality animatable avatars from text prompts. It follows a sequential Geometry->Texture->Animation paradigm, simplifying optimization through step-by-step generation. To tackle oversaturation, we introduce Adaptive Variational Parameter (AVP), representing avatars as an adaptive distribution during training. Additionally, we present Avatar-aware Score Distillation Sampling (ASDS), a novel technique that incorporates avatar-aware noise into rendered images for improved generation quality during optimization. Extensive evaluations confirm the superiority of X-Oscar over existing text-to-3D and text-to-avatar approaches. Our anonymous project page: https://xmu-xiaoma666.github.io/Projects/X-Oscar/.
@article{arxiv.2405.00954,
title = {X-Oscar: A Progressive Framework for High-quality Text-guided 3D Animatable Avatar Generation},
author = {Yiwei Ma and Zhekai Lin and Jiayi Ji and Yijun Fan and Xiaoshuai Sun and Rongrong Ji},
journal= {arXiv preprint arXiv:2405.00954},
year = {2024}
}