State of the Art on Diffusion Models for Visual Computing
Computer Graphics Forum · 2024-05 · 43 (2)
DOI: 10.1111/cgf.15063 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computer Graphics Forum, online erschienen am 30.04.2024. Verfasst von R. Po und 17 weiteren Autor:innen. Die Arbeit wurde bisher 114-mal zitiert (Crossref) und nennt 371 Literaturangaben. Gefördert von: Deutsche Forschungsgemeinschaft, Samsung, Stanford Institute for Human-Centered Artificial Intelligence, Stanford University.
Abstract (Originalsprache)
The field of visual computing is rapidly advancing due to the emergence of generative artificial intelligence (AI), which unlocks unprecedented capabilities for the generation, editing, and reconstruction of images, videos, and 3D scenes. In these domains, diffusion models are the generative AI architecture of choice. Within the last year alone, the literature on diffusion‐based tools and applications has seen exponential growth and relevant papers are published across the computer graphics, computer vision, and AI communities with new works appearing daily on arXiv. This rapid growth of the field makes it difficult to keep up with all recent developments. The goal of this state‐of‐the‐art report (STAR) is to introduce the basic mathematical concepts of diffusion models, implementation details and design choices of the popular Stable Diffusion model, as well as overview important aspects of these generative AI tools, including personalization, conditioning, inversion, among others. Moreover, we give a comprehensive overview of the rapidly growing literature on diffusion‐based generation and editing, categorized by the type of generated medium, including 2D images, videos, 3D objects, locomotion, and 4D scenes. Finally, we discuss available datasets, metrics, open challenges, and social implications. This STAR provides an intuitive starting point to explore this exciting topic for researchers, artists, and practitioners alike.
Förderung
- Deutsche Forschungsgemeinschaft
- Samsung
- Stanford Institute for Human-Centered Artificial Intelligence, Stanford University
Zitieren
APA
Po, R., Yifan, W., Golyanik, V., Aberman, K., Barron, J. T., Bermano, A., Chan, E., Dekel, T., Holynski, A., Kanazawa, A., Liu, C., Liu, L., Mildenhall, B., Nießner, M., Ommer, B., Theobalt, C., Wonka, P., & Wetzstein, G. (2024). State of the Art on Diffusion Models for Visual Computing. Computer Graphics Forum, 43(2). https://doi.org/10.1111/cgf.15063
Vancouver
Po R, Yifan W, Golyanik V, Aberman K, Barron JT, Bermano A, et al. State of the Art on Diffusion Models for Visual Computing. Computer Graphics Forum. 2024;43(2). doi:10.1111/cgf.15063
BibTeX
@article{Po2024,
author = {Po, R. and Yifan, W. and Golyanik, V. and Aberman, K. and Barron, J. T. and Bermano, A. and Chan, E. and Dekel, T. and Holynski, A. and Kanazawa, A. and Liu, C.K. and Liu, L. and Mildenhall, B. and Nießner, M. and Ommer, B. and Theobalt, C. and Wonka, P. and Wetzstein, G.},
title = {State of the Art on Diffusion Models for Visual Computing},
journal = {Computer Graphics Forum},
year = {2024},
volume = {43},
number = {2},
doi = {10.1111/cgf.15063}
}
EUVIMED