# Rodin: A Generative Model for Sculpting 3D Digital Avatars Using Diffusion

Source: [Rodin: A Generative Model for Sculpting 3D Digital Avatars Using Diffusion](<https://www.sota2.com/research/paper/68fa6e311e209131cea3e786>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-01T15:58:55.891Z

Authors: Tengfei Wang, Bo Zhang, Ting Zhang, Shuyang Gu, Jianmin Bao, Tadas Baltrusaitis, Jingjing Shen, Dong Chen, Fang Wen, Qifeng Chen, Baining Guo

Published: 2022-12-12T18:59:40.000Z

## Abstract

This paper presents a 3D generative model that uses diffusion models to automatically generate 3D digital avatars represented as neural radiance fields. A significant challenge in generating such avatars is that the memory and processing costs in 3D are prohibitive for producing the rich details required for high-quality avatars. To tackle this problem we propose the roll-out diffusion network (Rodin), which represents a neural radiance field as multiple 2D feature maps and rolls out these maps into a single 2D feature plane within which we perform 3D-aware diffusion. The Rodin model brings the much-needed computational efficiency while preserving the integrity of diffusion in 3D by using 3D-aware convolution that attends to projected features in the 2D feature plane according to their original relationship in 3D. We also use latent conditioning to orchestrate the feature generation for global coherence, leading to high-fidelity avatars and enabling their semantic editing based on text prompts. Finally, we use hierarchical synthesis to further enhance details. The 3D avatars generated by our model compare favorably with those produced by existing generative techniques. We can generate highly detailed avatars with realistic hairstyles and facial hair like beards. We also demonstrate 3D avatar generation from image or text as well as text-guided editability.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Unconditional 3D Avatar Generation on Synthetic 3D Avatar Dataset 100K avatars (test)](<https://www.sota2.com/research/sota/unconditional-3d-avatar-generation-on-synthetic-3d-avatar-dataset-100k-avatars-test.md>): Unconditional 3D Avatar Generation; Synthetic 3D Avatar Dataset 100K avatars (test); FID: 26.1.

- [Digital Avatar Creation on Synthetic Avatar](<https://www.sota2.com/research/sota/digital-avatar-creation-on-synthetic-avatar.md>): Digital Avatar Creation; Synthetic Avatar; PSNR: 18.8.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2212.06135>)

[PDF](<https://arxiv.org/pdf/2212.06135.pdf>)
