EUVIMEDEUVIMED entdecken
Crossref

MagicMan: Generative Novel View Synthesis of Humans with 3D-Aware Diffusion and Iterative Refinement

Xu He, Zhiyong Wu, Xiaoyu Li, Di Kang, Chaopeng Zhang, Jiangnan Ye, Liyang Chen, Xiangjun Gao, Han Zhang, Haolin Zhuang

Proceedings of the AAAI Conference on Artificial Intelligence · 11.04.2025 · 39 (3) · 3437-3445

DOI: 10.1609/aaai.v39i3.32356 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 11.04.2025. Verfasst von Xu He und 9 weiteren Autor:innen. Die Arbeit wurde bisher 26-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Existing works in single-image human reconstruction suffer from weak generalizability due to insufficient training data or 3D inconsistencies for a lack of comprehensive multi-view knowledge. In this paper, we introduce MagicMan, a human-specific multi-view diffusion model to generate high-quality novel views from a single reference image. As its core, we leverage a pre-trained 2D diffusion model as the generative prior for generalizability, with the parametric SMPL-X model as the 3D body prior to promote 3D awareness. To maintain consistency while generating denser views for improved 3D human reconstruction, we introduce hybrid multi-view attention to facilitate efficient and thorough information interchange across views. Besides, we present a geometry-aware dual branch to perform concurrent generation in both RGB and normal domains, further enhancing consistency via geometry cues. Last but not least, to address ill-shaped issues arising from inaccurate SMPL-X estimation, we propose a novel iterative refinement strategy, which progressively optimizes SMPL-X accuracy while enhancing the quality and consistency of the generated multi-views. Extensive experimental results demonstrate that our method significantly outperforms existing approaches in both novel view synthesis and subsequent 3D human reconstruction tasks.

Zitieren

APA

He, X., Wu, Z., Li, X., Kang, D., Zhang, C., Ye, J., Chen, L., Gao, X., Zhang, H., & Zhuang, H. (2025). MagicMan: Generative Novel View Synthesis of Humans with 3D-Aware Diffusion and Iterative Refinement. Proceedings of the AAAI Conference on Artificial Intelligence, 39(3), 3437-3445. https://doi.org/10.1609/aaai.v39i3.32356

Vancouver

He X, Wu Z, Li X, Kang D, Zhang C, Ye J, et al. MagicMan: Generative Novel View Synthesis of Humans with 3D-Aware Diffusion and Iterative Refinement. Proceedings of the AAAI Conference on Artificial Intelligence. 2025;39(3):3437-3445. doi:10.1609/aaai.v39i3.32356

BibTeX
@article{He2025,
  author = {He, Xu and Wu, Zhiyong and Li, Xiaoyu and Kang, Di and Zhang, Chaopeng and Ye, Jiangnan and Chen, Liyang and Gao, Xiangjun and Zhang, Han and Zhuang, Haolin},
  title = {MagicMan: Generative Novel View Synthesis of Humans with 3D-Aware Diffusion and Iterative Refinement},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2025},
  volume = {39},
  number = {3},
  pages = {3437--3445},
  doi = {10.1609/aaai.v39i3.32356}
}
Kostenlos starten →