PulseAugur
EN
LIVE 14:27:03

New methods generate realistic 3D avatars from single photos

By PulseAugur Editorial · [3 sources] ·

Researchers have developed two new methods for generating realistic 3D human avatars from single images. Splatshot combines 3D Gaussian Splatting with diffusion models to ensure both geometric consistency and photorealism, even with unconstrained input photos. HumanNOVA utilizes a large-scale data generation pipeline and a feed-forward architecture to create avatars rapidly and with high fidelity, requiring no test-time optimization. AI

IMPACT Advances in single-image 3D avatar generation could accelerate applications in virtual reality, gaming, and personalized digital experiences.

RANK_REASON Two research papers published on arXiv detailing new methods for 3D avatar generation.

Read on arXiv cs.CV →

AI-generated summary · Google Gemini · from 3 sources. How we write summaries →

COVERAGE [3]

  1. arXiv cs.CV TIER_1 English(EN) · Hao Liang, Zhixuan Ge, Soumendu Majee, Joanna Li, Ashok Veeraraghavan, Guha Balakrishnan ·

    Splatshot: 3D Face Avatar Generation from a Single Unconstrained Photo

    arXiv:2606.01493v1 Announce Type: new Abstract: Reconstructing a photorealistic 3D face avatar from a single unconstrained photograph is challenging: feed-forward 3D Gaussian Splatting (3DGS) models degrade on out-of-distribution inputs, while pretrained diffusion models produce …

  2. arXiv cs.CV TIER_1 English(EN) · Hezhen Hu, Wangbo Zhao, Lanqing Guo, Hanwen Jiang, Jonathan C. Liu, Zhiwen Fan, Kai Wang, Zhangyang Wang, Georgios Pavlakos ·

    HumanNOVA: Photorealistic, Universal and Rapid 3D Human Avatar Modeling from a Single Image

    arXiv:2606.02573v1 Announce Type: new Abstract: In this paper, we present HumanNOVA, a photorealistic, universal, and rapid model for generating 3D human avatars from a single RGB image. Achieving both photorealism and generalization is challenging due to the scarcity of diverse,…

  3. arXiv cs.CV TIER_1 English(EN) · Georgios Pavlakos ·

    HumanNOVA: Photorealistic, Universal and Rapid 3D Human Avatar Modeling from a Single Image

    In this paper, we present HumanNOVA, a photorealistic, universal, and rapid model for generating 3D human avatars from a single RGB image. Achieving both photorealism and generalization is challenging due to the scarcity of diverse, high-quality 3D human data. To address this, we…

RELATED ENTITIES

RELATED TOPICS