CoChat
Sign Up
Cited by 338Generative Adversarial Networks and Image SynthesisImage and Signal Denoising MethodsAdversarial Robustness in Machine Learning

Diffusion Autoencoders: Toward a Meaningful and Decodable Representation

Open access
Preechakul, Konpat, Chatthee, Nattanat, Wizadwongsa, Suttisak, Suwajanakorn, Supasorn·2022·2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
DOI 10.1109/cvpr52688.2022.01036
Open this paper in CoChat — free accountOpen PDF

Abstract

Diffusion probabilistic models (DPMs) have achieved remarkable quality in image generation that rivals GANs'. But unlike GANs, DPMs use a set of latent variables that lack semantic meaning and cannot serve as a useful representation for other tasks. This paper explores the possibility of using DPMs for representation learning and seeks to extract a meaningful and decodable representation of an input image via autoencoding. Our key idea is to use a learnable encoder for discovering the high-level semantics, and a DPM as the decoder for modeling the remaining stochastic variations. Our method can encode any image into a two-part latent code where the first part is semantically meaningful and linear, and the second part captures stochastic details, allowing near-exact reconstruction. This capability enables challenging applications that currently foil GAN-based methods, such as attribute manipulation on real images. We also show that this two-level encoding improves denoising efficiency and naturally facilitates various downstream tasks including few-shot conditional sampling. Please visit our page: https://Diff-AE.github.io/

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/cvpr52688.2022.01036

Cites

9 of 9 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Preechakul et al. explore the potential of utilizing Diffusion Probability Models for representation learning, encoding any image into a two-part latent code (Diff-AE).

    “Furthermore, Preechakul et al. [52] explore the potential of utilizing Difusion Probability Models (DPMs) [19] for representation learning.”

    From DEEPFAKER: A Unified Evaluation Platform for Facial Deepfake and Detection Models (2024) · cites this paper as [52]

  • The BlendAnything design combines the factorized representation of Diffusion Autoencoders (DiffAE) with the any-topology motion generation setting of AnyTop.

    “The design combines the factorized representation of Diffusion Autoencoders (DiffAE) [Preechakul et al. 2022] with the any-topology motion generation setting of AnyTop [Gat et al. 2025].”

    From BlendAnything: A Blender Plugin for Cross-Topology Motion Blending (2026) · cites this paper as Preechakul et al. 2022

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 45

  1. [1][1] Rameen Abdal, Yipeng Qin, and Peter Wonka. Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space? pages 4432-4441, 2019. -
  2. [2][2] Rameen Abdal, Yipeng Qin, and Peter Wonka. Image2StyleGAN++: How to Edit the Embedded Images? pages 8296-8305, 2020. -
  3. [3][3] Rameen Abdal, Peihao Zhu, Niloy Mitra, and Peter Wonka. StyleFlow: Attribute-conditioned Exploration of StyleGANGenerated Images using Conditional Continuous Normalizing Flows. ACMTransactions on Graphics , 40(3):1-21, May 2021. arXiv: 2008.02401. - · arXiv
  4. [4][4] Guillaume Alain and Yoshua Bengio. Understanding intermediate layers using linear classifier probes. arXiv:1610.01644 [cs, stat] , Nov. 2018. arXiv: 1610.01644. - · arXiv
  5. [6][6] Nanxin Chen, Yu Zhang, Heiga Zen, Ron J. Weiss, Mohammad Norouzi, and William Chan. WaveGrad: Estimating Gradients for Waveform Generation. arXiv:2009.00713 [cs, eess, stat] , Oct. 2020. arXiv: 2009.00713. - · arXiv
  6. [7][7] Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. A Simple Framework for Contrastive Learning of Visual Representations. In Proceedings of the 37th International Conference on Machine Learning , pages 15971607. PMLR, Nov. 2020. ISSN: 2640-3498. -
  7. [8][8] Jooyoung Choi, Sungwon Kim, Yonghyun Jeong, Youngjune Gwon, and Sungroh Yoon. ILVR: Conditioning Method for Denoising Diffusion Probabilistic Models. arXiv:2108.02938 [cs] , Sept. 2021. arXiv: 2108.02938. - · arXiv
  8. [9][9] Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. ImageNet: A large-scale hierarchical image database. In 2009 IEEE Conference on Computer Vision and Pattern Recognition , pages 248-255, June 2009. ISSN: 1063-6919. -
  9. [10][10] Jiankang Deng, Jia Guo, Niannan Xue, and Stefanos Zafeiriou. ArcFace: Additive Angular Margin Loss for Deep Face Recognition. pages 4690-4699, 2019. -
  10. [13][13] Patrick Esser, Robin Rombach, and Bjorn Ommer. Taming Transformers for High-Resolution Image Synthesis. pages 12873-12883, 2021. -
  11. [14][14] Daniel Freeman, David Ha, and Luke Metz. Learning to Predict Without Looking Ahead: World Models Without Forward Prediction. In Advances in Neural Information Processing Systems , volume 32. Curran Associates, Inc., 2019. -
  12. [15][15] Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative Adversarial Nets. In Advances in 16. Neural Information Processing Systems , volume 27. Curran Associates, Inc., 2014. -
  13. [16][16] Jean-Bastien Grill, Florian Strub, Florent Altch´ e, Corentin Tallec, Pierre Richemond, Elena Buchatskaya, Carl Doersch, Bernardo Avila Pires, Zhaohan Guo, Mohammad Gheshlaghi Azar, Bilal Piot, koray kavukcuoglu, Remi Munos, and Michal Valko. Bootstrap Your Own Latent - A New Approach to Self-Supervised Learning. In H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems , volume 33, pages 21271-21284. Curran Associates, Inc., 2020. -
  14. [17][17] David Ha and J¨ urgen Schmidhuber. Recurrent World Models Facilitate Policy Evolution. In S. Bengio, H. Wallach, H. Larochelle, K. Grauman, N. Cesa-Bianchi, and R. Garnett, editors, Advances in Neural Information Processing Systems , volume 31. Curran Associates, Inc., 2018. -
  15. [19][19] Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum Contrast for Unsupervised Visual Representation Learning. pages 9729-9738, 2020. -
  16. [22][22] Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising Diffusion Probabilistic Models. In H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems , volume 33, pages 6840-6851. Curran Associates, Inc., 2020. -
  17. [23][23] Jonathan Ho, Chitwan Saharia, William Chan, David J. Fleet, Mohammad Norouzi, and Tim Salimans. Cascaded Diffusion Models for High Fidelity Image Generation. Journal of Machine Learning Research , 23(47):1-33, 2022. -
  18. [24][24] Erik H¨ ark¨ onen, Aaron Hertzmann, Jaakko Lehtinen, and Sylvain Paris. GANSpace: Discovering Interpretable GAN Controls. In H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems , volume 33, pages 9841-9850. Curran Associates, Inc., 2020. -
  19. [25][25] Alexia Jolicoeur-Martineau, Ke Li, R´ emi Pich´ e-Taillefer, Tal Kachman, and Ioannis Mitliagkas. Gotta Go Fast When Generating Data with Score-Based Models. arXiv:2105.14080 [cs, math, stat] , May 2021. arXiv: 2105.14080. - · arXiv
  20. [26][26] Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. Progressive Growing of GANs for Improved Quality, Stability, and Variation. arXiv:1710.10196 [cs, stat] , Feb. 2018. arXiv: 1710.10196. - · arXiv
  21. [27][27] Tero Karras, Samuli Laine, and Timo Aila. A StyleBased Generator Architecture for Generative Adversarial Networks. pages 4401-4410, 2019. -
  22. [28][28] Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Analyzing and Improving the Image Quality of StyleGAN. pages 8110-8119, 2020. -
  23. [31][31] Max W. Y. Lam, Jun Wang, Rongjie Huang, Dan Su, and Dong Yu. Bilateral Denoising Diffusion Models. arXiv:2108.11514 [cs, eess] , Sept. 2021. arXiv: 2108.11514. - · arXiv
  24. [33][33] Ziwei Liu, Ping Luo, Xiaogang Wang, and Xiaoou Tang. Deep Learning Face Attributes in the Wild. In 2015 IEEE International Conference on Computer Vision (ICCV) , pages 3730-3738, Dec. 2015. ISSN: 2380-7504. -
  25. [34][34] Shitong Luo and Wei Hu. Diffusion Probabilistic Models for 3D Point Cloud Generation. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 2836-2844, Nashville, TN, USA, June 2021. IEEE. -
  26. [36][36] Alexander Quinn Nichol and Prafulla Dhariwal. Improved Denoising Diffusion Probabilistic Models. In Proceedings of the 38th International Conference on Machine Learning , pages 8162-8171. PMLR, July 2021. ISSN: 2640-3498. -
  27. [37][37] Or Patashnik, Zongze Wu, Eli Shechtman, Daniel Cohen-Or, and Dani Lischinski. StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery. pages 2085-2094, 2021. -
  28. [38][38] Ali Razavi, Aaron van den Oord, and Oriol Vinyals. Generating Diverse High-Fidelity Images with VQ-VAE-2. In H. Wallach, H. Larochelle, A. Beygelzimer, F. d' Alch´ e-Buc, E. Fox, and R. Garnett, editors, Advances in Neural Information Processing Systems , volume 32. Curran Associates, Inc., 2019. -
  29. [39][39] Elad Richardson, Yuval Alaluf, Or Patashnik, Yotam Nitzan, Yaniv Azar, Stav Shapiro, and Daniel Cohen-Or. Encoding in Style: A StyleGAN Encoder for Image-to-Image Translation. pages 2287-2296, 2021. -
  30. [40][40] Olaf Ronneberger, Philipp Fischer, and Thomas Brox. UNet: Convolutional Networks for Biomedical Image Segmentation. In Nassir Navab, Joachim Hornegger, William M. Wells, and Alejandro F. Frangi, editors, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015 , Lecture Notes in Computer Science, pages 234-241, Cham, 2015. Springer International Publishing. -
  31. [41][41] Mihaela Rosca, Balaji Lakshminarayanan, and Shakir Mohamed. Distribution Matching in Variational Inference. arXiv:1802.06847 [cs, stat] , June 2019. arXiv: 1802.06847. - · arXiv
  32. [42][42] Oleh Rybkin, Kostas Daniilidis, and Sergey Levine. Simple and Effective VAE Training with Calibrated Decoders. 16. In Proceedings of the 38th International Conference on Machine Learning , pages 9179-9189. PMLR, July 2021. ISSN: 2640-3498. -
  33. [43][43] Chitwan Saharia, Jonathan Ho, William Chan, Tim Salimans, David J. Fleet, and Mohammad Norouzi. Image Super-Resolution via Iterative Refinement. arXiv:2104.07636 [cs, eess] , June 2021. arXiv: 2104.07636. - · arXiv
  34. [44][44] Yujun Shen, Jinjin Gu, Xiaoou Tang, and Bolei Zhou. Interpreting the Latent Space of GANs for Semantic Face Editing. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 9240-9249, June 2020. ISSN: 2575-7075. -
  35. [46][46] Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep Unsupervised Learning using Nonequilibrium Thermodynamics. In Proceedings of the 32nd International Conference on Machine Learning , pages 2256-2265. PMLR, June 2015. ISSN: 1938-7228. -
  36. [48][48] Yang Song and Stefano Ermon. Generative Modeling by Estimating Gradients of the Data Distribution. In H. Wallach, H. Larochelle, A. Beygelzimer, F. d' Alch´ e-Buc, E. Fox, and R. Garnett, editors, Advances in Neural Information Processing Systems , volume 32. Curran Associates, Inc., 2019. -
  37. [50][50] Arash Vahdat and Jan Kautz. NVAE: A Deep Hierarchical Variational Autoencoder. In H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems , volume 33, pages 19667-19679. Curran Associates, Inc., 2020. -
  38. [51][51] Aaron van den Oord, Nal Kalchbrenner, Lasse Espeholt, koray kavukcuoglu, Oriol Vinyals, and Alex Graves. Conditional Image Generation with PixelCNN Decoders. In D. Lee, M. Sugiyama, U. Luxburg, I. Guyon, and R. Garnett, editors, Advances in Neural Information Processing Systems , volume 29. Curran Associates, Inc., 2016. -
  39. [52][52] Aaron van den Oord, Oriol Vinyals, and koray kavukcuoglu. Neural Discrete Representation Learning. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems , volume 30. Curran Associates, Inc., 2017. -
  40. [55][55] Po-Wei Wu, Yu-Jing Lin, Che-Han Chang, Edward Y. Chang, and Shih-Wei Liao. RelGAN: Multi-Domain Imageto-Image Translation via Relative Attributes. pages 59145922, 2019. -
  41. [56][56] Yuxin Wu and Kaiming He. Group Normalization. pages 3-19, 2018. -
  42. [57][57] Zongze Wu, Dani Lischinski, and Eli Shechtman. StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation. pages 12863-12872, 2021. -
  43. [58][58] Weihao Xia, Yulun Zhang, Yujiu Yang, Jing-Hao Xue, Bolei Zhou, and Ming-Hsuan Yang. GAN Inversion: A Survey. arXiv:2101.05278 [cs] , Aug. 2021. arXiv: 2101.05278. - · arXiv
  44. [60][60] Fisher Yu, Ari Seff, Yinda Zhang, Shuran Song, Thomas Funkhouser, and Jianxiong Xiao. LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop. arXiv:1506.03365 [cs] , June 2016. arXiv: 1506.03365. - · arXiv
  45. [61][61] Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shechtman, and Oliver Wang. The Unreasonable Effectiveness of Deep Features as a Perceptual Metric. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 586-595, Salt Lake City, UT, June 2018. IEEE.

Cited by

Sorted by citations

Authors
Preechakul, K., Chatthee, N., Wizadwongsa, S., & Suwajanakorn, S.
Affiliations
Vidyasirimedhi Institute of Science and Technology (TH)
ORCID
Venue
2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
Publisher
IEEE
Published
2022-06
Pages
10609-10619
ISSN
1063-6919 (ISSN-L)
DOI
10.1109/cvpr52688.2022.01036
OpenAlex
W3217030260
Cited by
338
References
9
Citations by year
2026: 57 · 2025: 112 · 2024: 113 · 2023: 55 · 2022: 1
FWCI
25.48
Type
conference paper
Language
en
Open access
Licence
Locations
  • IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Generative Adversarial Networks and Image Synthesis
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Signal Denoising Methods
  • Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
Keywords
diffusion probabilistic models, diffusion autoencoders, representation learning, image denoising, image generation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review