English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent advances in 3D avatar generation have gained significant attentions.These breakthroughs aim to produce more realistic animatable avatars, narrowingthe gap between virtual and real-world experiences. Most of existing worksemploy Score Distillation Sampling (SDS) loss, combined with a differentiablerenderer and text condition, to guide a diffusion model in generating 3Davatars. However, SDS often generates oversmoothed results with few facialdetails, thereby lacking the diversity compared with ancestral sampling. On theother hand, other works generate 3D avatar from a single image, where thechallenges of unwanted lighting effects, perspective views, and inferior imagequality make them difficult to reliably reconstruct the 3D face meshes with thealigned complete textures. In this paper, we propose a novel 3D avatargeneration approach termed UltrAvatar with enhanced fidelity of geometry, andsuperior quality of physically based rendering (PBR) textures without unwantedlighting. To this end, the proposed approach presents a diffuse colorextraction model and an authenticity guided texture diffusion model. The formerremoves the unwanted lighting effects to reveal true diffuse colors so that thegenerated avatars can be rendered under various lighting conditions. The latterfollows two gradient-based guidances for generating PBR textures to renderdiverse face-identity features and details better aligning with 3D meshgeometry. We demonstrate the effectiveness and robustness of the proposedmethod, outperforming the state-of-the-art methods by a large margin in theexperiments.

UltrAvatar: A Realistic Animatable 3D Avatar Diffusion Model with Authenticity Guided Textures