StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets
Open accessThe DOI matches its Crossref record; 28 of the 30 references it cites were found in a registry or an index, and 2 are not checked yet.
Abstract
Computer graphics has experienced a recent surge of data-centric approaches for photorealistic and controllable content creation. StyleGAN in particular sets new standards for generative modeling regarding image quality and controllability. However, StyleGAN’s performance severely degrades on large unstructured datasets such as ImageNet. StyleGAN was designed for controllability; hence, prior works suspect its restrictive design to be unsuitable for diverse datasets. In contrast, we find the main limiting factor to be the current training strategy. Following the recently introduced Projected GAN paradigm, we leverage powerful neural network priors and a progressive growing strategy to successfully train the latest StyleGAN3 generator on ImageNet. Our final model, StyleGAN-XL, sets a new state-of-the-art on large-scale image synthesis and is the first to generate images at a resolution of 10242 at such a dataset scale. We demonstrate that this model can invert and edit images beyond the narrow domain of portraits or specific object classes. Code, models, and supplementary videos can be found at https://sites.google.com/view/stylegan-xl/ .
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“StyleGAN-XL overcomes the scalability issues of previous models, offering better sample diversity, generalization, and image fidelity.”
From Deepfake Media Forensics: Status and Future Challenges (2025) · cites this paper as [36]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Our hypothesis could explain the phenomenon that conditional generation is easier than unconditional (Mirza & Osindero, 2014;Liu et al., 2020;Sauer et al., 2022), as the data we condition on may have the same platonic structure as the data we are generating.”
From The Platonic Representation Hypothesis (2024) · cites this paper as Sauer et al., 2022
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“StyleGAN-XL [36] is an extension of the StyleGAN family, designed to generate high-quality images on diverse and unstructured datasets, particularly ImageNet.”
From Deepfake Media Forensics: Status and Future Challenges (2025) · cites this paper as [36]
The cited source says what the claim attributes to it.
In the citing paper
“The model achieves state-of-the-art performance in large-scale image synthesis, generating high-resolution images up to 1024 × 1024.”
From Deepfake Media Forensics: Status and Future Challenges (2025) · cites this paper as [36]
The cited source says what the claim attributes to it.
In the citing paper
“StyleGAN-XL incorporates Projected GAN training and a progressive growing strategy to enhance stability and scalability.”
From Deepfake Media Forensics: Status and Future Challenges (2025) · cites this paper as [36]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333 · cited 1×
“However, when trained on large and unstructured datasets like ImageNet [Deng et al. 2009], StyleGANs do not achieve satisfactory results yet.”
MethodGoing deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
The unreasonable effectiveness of deep features as a perceptual metric
Zhang, R., Isola, P., Efros, A. A., Shechtman, E., & Wang, O. (2018) · arXiv · doi:10.48550/arxiv.1801.03924 · Cited by 24,812
The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
Zhang, R., Isola, P., Efros, A. A., Shechtman, E., & Wang, O. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00068 · Cited by 14,623
Unsupervised representation learning with deep convolutional generative adversarial networks
Radford, A., Metz, L., & Chintala, S. (2015) · arXiv · doi:10.48550/arxiv.1511.06434 · Cited by 12,235 · cited 2×
“However, its architecture is more restrictive than a standard generator network [Karras et al. 2018; Radford et al. 2016] which seems to come at a price when training on complex”
BackgroundA Style-Based Generator Architecture for Generative Adversarial Networks
Karras, T., Laine, S., & Aila, T. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00453 · Cited by 11,714 · cited 7×
“Regularization improves results on uni-modal datasets like FFHQ [27] or LSUN [63], whereas it can be detrimental on multi-modal datasets [5, 19].”
BackgroundInfluentialProceedings of the 25th international conference on Machine learning - ICML '08
(2008) · doi:10.1145/1390156 · Cited by 6,680
Analyzing and Improving the Image Quality of StyleGAN
Karras, T., Laine, S., Aittala, M., Hellsten, J., Lehtinen, J., & Aila, T. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00813 · Cited by 6,332
Analyzing and Improving the Image Quality of StyleGAN
Karras, T., Laine, S., Aittala, M., Hellsten, J., Lehtinen, J., & Aila, T. (2019) · arXiv · doi:10.48550/arxiv.1912.04958 · Cited by 6,332
SwinIR: Image Restoration Using Swin Transformer
Liang, J., Cao, J., Sun, G., Zhang, K., Van Gool, L., & Timofte, R. (2021) · 2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) · doi:10.1109/iccvw54120.2021.00210 · Cited by 4,734 · cited 2×
“the data with a super-resolution network [Liang et al. 2021], see supplementary.”
Method
Research works printed but not linked: 46
- Yuval Alaluf, Or Patashnik, Zongze Wu, Asif Zamir, Eli Shechtman, Dani Lischinski, and Daniel Cohen-Or. 2022. Third Time's the Charm? Image and Video Editing with StyleGAN3. arXiv.org abs/2201.13433 (2022). https://arxiv.org/abs/2201.13433 · arXiv
- Yuval Alaluf, Or Patashnik, and Daniel Cohen-Or. 2021. ReStyle: A Residual-Based StyleGAN Encoder via Iterative Refinement. Proc. of the IEEE International Conf. on Computer Vision (ICCV) abs/2104.02699 (2021). https://arxiv.org/abs/2104.02699 · arXiv
- Andrew Brock, Jeff Donahue, and Karen Simonyan. 2019. Large Scale GAN Training for High Fidelity Natural Image Synthesis. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https://openreview.net/forum? id=B1xsqj09Fm
- Arantxa Casanova, Marlène Careil, Jakob Verbeek, Michal Drozdzal, and Adriana Romero-Soriano. 2021. Instance-Conditioned GAN. In Advances in Neural Information Processing Systems (NeurIPS) .
- Lucy Chai, Jun-Yan Zhu, Eli Shechtman, Phillip Isola, and Richard Zhang. 2021. Ensembling With Deep Generative Views. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) . Computer Vision Foundation / IEEE, 1499715007. https://openaccess.thecvf.com/content/CVPR2021/html/Chai\_Ensembling\_ With\_Deep\_Generative\_Views\_CVPR\_2021\_paper.html momentum contrastive learning. arxiv.org:2003.04297
- Ting-Yun Chang and Chi-Jen Lu. 2020. TinyGAN: Distilling BigGAN for Conditional Image Generation. In Proc. of the Asian Conf. on Computer Vision (ACCV) (Lecture Notes in Computer Science, Vol. 12625) , Hiroshi Ishikawa, Cheng-Lin Liu, Tomás Pajdla, and Jianbo Shi (Eds.). 509-525. https://doi.org/10.1007/978-3-030-69538-5\_31 Xinlei Chen, Haoqi Fan, and Ross Girshick Kaiming He. 2020. Improved baselines with · DOI
- Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. 2021. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https://openreview.net/ forum?id=YicbFdNTTy
- Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron C. Courville, and Yoshua Bengio. 2014. Generative Adversarial Nets. In Advances in Neural Information Processing Systems (NeurIPS) ,
- Tom George Grigg, Dan Busbridge, Jason Ramapuram, and Russ Webb. 2021. Do Self-Supervised and Supervised Methods Learn Similar Visual Representations? arxiv.org:2110.00528
- Timofey Grigoryev, Andrey Voynov, and Artem Babenko. 2022. When, Why, and Which Pretrained GANs Are Useful?. In Proc. of the International Conf. on Learning Representations (ICLR) . https://openreview.net/forum?id=4Ycr8oeCoIh
- Gwern. 2020. Making Anime Faces with StyleGAN . Retrieved January 16, 2022 from https://www.gwern.net/Faces#stylegan2-ext-modifications/
- Erik Härkönen, Aaron Hertzmann, Jaakko Lehtinen, and Sylvain Paris. 2020. GANSpace: Discovering Interpretable GAN Controls. In Advances in Neural Information Processing Systems (NeurIPS) . https://proceedings.neurips.cc/paper/2020/hash/ 6fe43269967adbb64ec6149852b5cc3e-Abstract.html
- Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. 2017. GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium. In Advances in Neural Information Processing Systems (NeurIPS) . 6626-6637. https://proceedings.neurips.cc/paper/2017/hash/ 8a1d694707eb0fefe65871369074926d-Abstract.html
- Ali Jahanian, Lucy Chai, and Phillip Isola. 2020. On the "steerability" of generative adversarial networks. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https://openreview.net/forum?id=HylsTT4FvB
- Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. 2018. Progressive Growing of GANs for Improved Quality, Stability, and Variation. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https://openreview.net/ forum?id=Hk99zCeAb
- Tero Karras, Miika Aittala, Samuli Laine, Erik Härkönen, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. 2021. Alias-Free Generative Adversarial Networks. In Advances in Neural Information Processing Systems (NeurIPS) .
- Umut Kocasari, Alara Dirik, Mert Tiftikci, and Pinar Yanardag. 2022. StyleMC: MultiChannel Based Fast Text-Guided Image Generation and Manipulation. Proc. of the IEEE Winter Conference on Applications of Computer Vision (WACV) (2022). https: //arxiv.org/abs/2112.08493
- Diederik P. Kingma and Jimmy Ba. 2015. Adam: A Method for Stochastic Optimization. In Proc. of the International Conf. on Learning Representations (ICLR) .
- Daiqing Li, Huan Ling, Seung Wook Kim, Karsten Kreis, Adela Barriuso, Sanja Fidler, and Antonio Torralba. 2022. BigDatasetGAN: Synthesizing ImageNet with Pixelwise Annotations. arXiv.org (2022). https://arxiv.org/abs/2201.04684 · arXiv
- Huan Ling, Karsten Kreis, Daiqing Li, Seung Wook Kim, Antonio Torralba, and Sanja Fidler. 2021. EditGAN: High-Precision Semantic Image Editing. arXiv.org abs/2111.03186 (2021). https://arxiv.org/abs/2111.03186 · arXiv
- Bingchen Liu, Yizhe Zhu, Kunpeng Song, and Ahmed Elgammal. 2021. Towards Faster and Stabilized GAN Training for High-fidelity Few-shot Image Synthesis. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https: //openreview.net/forum?id=1Fqg133qRaI
- Lars M. Mescheder, Andreas Geiger, and Sebastian Nowozin. 2018. Which Training Methods for GANs do actually Converge?. In Proc. of the International Conf. on Machine learning (ICML) (Proceedings of Machine Learning Research, Vol. 80) . PMLR, 3478-3487. http://proceedings.mlr.press/v80/mescheder18a.html
- Takeru Miyato, Toshiki Kataoka, Masanori Koyama, and Yuichi Yoshida. 2018. Spectral Normalization for Generative Adversarial Networks. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https://openreview.net/ forum?id=B1QRgziT-
- Alexander Quinn Nichol and Prafulla Dhariwal. 2021. Improved Denoising Diffusion Probabilistic Models. In Proc. of the International Conf. on Machine learning (ICML) (Proceedings of Machine Learning Research, Vol. 139) . PMLR, 8162-8171. http:// proceedings.mlr.press/v139/nichol21a.html
- Or Patashnik, Zongze Wu, Eli Shechtman, Daniel Cohen-Or, and Dani Lischinski. 2021. StyleCLIP: Text-Driven Manipulation of StyleGAN Imagery. In Proc. of the IEEE International Conf. on Computer Vision (ICCV) . 2085-2094.
- Etienne Perot, Pierre de Tournemire, Davide Nitti, Jonathan Masci, and Amos Sironi. 2020. Learning to Detect Objects with a 1 Megapixel Event Camera. In Advances in Neural Information Processing Systems (NeurIPS) . https://proceedings.neurips.cc/ paper/2020/hash/c213877427b46fa96cff6c39e837ccee-Abstract.html
- Phillip Pope, Chen Zhu, Ahmed Abdelkader, Micah Goldblum, and Tom Goldstein. 2021. The Intrinsic Dimension of Images and Its Impact on Learning. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https: //openreview.net/forum?id=XJk19XzGq2J
- Maithra Raghu, Thomas Unterthiner, Simon Kornblith, Chiyuan Zhang, and Alexey Dosovitskiy. 2021. Do Vision Transformers See Like Convolutional Neural Networks?. In Advances in Neural Information Processing Systems (NeurIPS) .
- Daniel Roich, Ron Mokady, Amit H. Bermano, and Daniel Cohen-Or. 2021. Pivotal Tuning for Latent-based Editing of Real Images. arXiv.org (2021). https://arxiv.org/ abs/2106.05744 · arXiv
- Chitwan Saharia, Jonathan Ho, William Chan, Tim Salimans, David J. Fleet, and Mohammad Norouzi. 2021. Image Super-Resolution via Iterative Refinement. arxiv.org:2104.07636
- Tim Salimans, Ian J. Goodfellow, Wojciech Zaremba, Vicki Cheung, Alec Radford, and Xi Chen. 2016. Improved Techniques for Training GANs. In Advances in Neural Information Processing Systems (NeurIPS) . 2226-2234. https://proceedings.neurips. cc/paper/2016/hash/8a3363abe792db2d8761d6403605aeb7-Abstract.html
- Axel Sauer, Kashyap Chitta, Jens Müller, and Andreas Geiger. 2021. Projected GANs Converge Faster. In Advances in Neural Information Processing Systems (NeurIPS) .
- Axel Sauer and Andreas Geiger. 2021. Counterfactual Generative Networks. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https: //openreview.net/forum?id=BXewfAYMmJw
- Yujun Shen and Bolei Zhou. 2021. Closed-Form Factorization of Latent Semantics in GANs. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) . 1532-1540. https://openaccess.thecvf.com/content/CVPR2021/html/Shen\_Closed-Form\_Factorization\_of\_Latent\_Semantics\_in\_GANs\_CVPR\_2021\_paper.html
- Jiaming Song, Chenlin Meng, and Stefano Ermon. 2021. Denoising Diffusion Implicit Models. In Proc. of the International Conf. on Learning Representations (ICLR) . https: //openreview.net/forum?id=St1giarCHLP
- Nurit Spingarn, Ron Banner, and Tomer Michaeli. 2021. GAN "Steerability" without optimization. In Proc. of the International Conf. on Learning Representations (ICLR) . OpenReview.net. https://openreview.net/forum?id=zDy\_nQCXiIj
- Mingxing Tan and Quoc V. Le. 2019. EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. In Proc. of the International Conf. on Machine learning (ICML) (Proceedings of Machine Learning Research, Vol. 97) , Kamalika Chaudhuri and
- Matthew Tancik, Pratul P. Srinivasan, Ben Mildenhall, Sara Fridovich-Keil, Nithin Raghavan, Utkarsh Singhal, Ravi Ramamoorthi, Jonathan T. Barron, and Ren Ng. 2020. Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains. In Advances in Neural Information Processing Systems (NeurIPS) . https://proceedings.neurips.cc/paper/2020/hash/ 55053683268957697aa39fba6f231c68-Abstract.html
- Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, and Hervé Jégou. 2021. Training data-efficient image transformers & distillation through attention. In Proc. of the International Conf. on Machine learning (ICML) (Proceedings of Machine Learning Research, Vol. 139) . PMLR, 10347-10357. http://proceedings.mlr.press/v139/touvron21a.html
- Aäron van den Oord, Oriol Vinyals, and Koray Kavukcuoglu. 2017. Neural Discrete Representation Learning. In Advances in Neural Information Processing Systems (NeurIPS) , Isabelle Guyon, Ulrike von Luxburg, Samy Bengio, Hanna M. Wallach, Rob Fergus, S. V. N. Vishwanathan, and Roman Garnett (Eds.). 6306-6315. https://proceedings. neurips.cc/paper/2017/hash/7a98af17e63a0ac09ce2e96d03992fbc-Abstract.html
- Daniel Watson, Jonathan Ho, Mohammad Norouzi, and William Chan. 2021. Learning to Efficiently Sample from Diffusion Probabilistic Models. CoRR abs/2106.03802 (2021). https://arxiv.org/abs/2106.03802 · arXiv
- Andrey Voynov and Artem Babenko. 2020. Unsupervised Discovery of Interpretable Directions in the GAN Latent Space. In Proc. of the International Conf. on Machine learning (ICML) . PMLR, 9786-9796. http://proceedings.mlr.press/v119/voynov20a. html
- Zongze Wu, Dani Lischinski, and Eli Shechtman. 2021. StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) . Computer Vision Foundation / IEEE, 12863-12872. https://openaccess.thecvf.com/content/CVPR2021/html/Wu\_ StyleSpace\_Analysis\_Disentangled\_Controls\_for\_StyleGAN\_Image\_Generation\_ CVPR\_2021\_paper.html
- Rui Xu, Xintao Wang, Kai Chen, Bolei Zhou, and Chen Change Loy. 2021. Positional Encoding As Spatial Inductive Bias in GANs. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR) . Computer Vision Foundation / IEEE, 1356913578. https://openaccess.thecvf.com/content/CVPR2021/html/Xu\_Positional\_ Encoding\_As\_Spatial\_Inductive\_Bias\_in\_GANs\_CVPR\_2021\_paper.html
- Fisher Yu, Yinda Zhang, Shuran Song, Ari Seff, and Jianxiong Xiao. 2015. LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop. arxiv.org:1506.03365
- Shengyu Zhao, Zhijian Liu, Ji Lin, Jun-Yan Zhu, and Song Han. 2020. Differentiable Augmentation for Data-Efficient GAN Training. In Advances in Neural Information Processing Systems (NeurIPS) . https://proceedings.neurips.cc/paper/2020/hash/ 55479c55ebd1efd3ff125f1337100388-Abstract.html
Cited by
Scalable Diffusion Models with Transformers
Peebles, W., & Xie, S. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00387 · Cited by 2,284
“Our method achieves the lowest FID of all prior generative models, including the previous state-of-the-art StyleGANXL [50].”
MethodScalable Diffusion Models with Transformers
Peebles, W., & Xie, S. (2022) · arXiv · doi:10.48550/arxiv.2212.09748 · Cited by 2,284
“Our method achieves the lowest FID of all prior generative models, including the previous state-of-the-art StyleGANXL [50].”
MethodVisual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction
Tian, K., Jiang, Y., Yuan, Z., Peng, B., & Wang, L. (2024) · neural information processing systems · doi:10.52202/079017-2694 · Cited by 1,209
SiT: Exploring Flow and Diffusion-Based Generative Models with Scalable Interpolant Transformers
Ma, N., Goldstein, M., Albergo, M. S., Boffi, N. M., Vanden-Eijnden, E., & Xie, S. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72980-5_2 · Cited by 921
Improved Distribution Matching Distillation for Fast Image Synthesis
Yin, T., Gharbi, M., Park, T., Zhang, R., Shechtman, E., Durand, F., & Freeman, W. T. (2024) · neural information processing systems · doi:10.52202/079017-1505 · Cited by 873
Align Your Latents: High-Resolution Video Synthesis with Latent Diffusion Models
Blattmann, A., Rombach, R., Ling, H., Dockhorn, T., Kim, S. W., Fidler, S., & Kreis, K. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.02161 · Cited by 724
“The most powerful models today are built on generative adversarial networks [21, 38–40, 75 ], autoregressive transformers [ To this end, we build on latent diffusion models (LDMs), which can reduce the heavy computational burden when training on high-resolution images [65].”
BackgroundRepresentation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think
Yu, S., Kwak, S., Jang, H., Jeong, J., Huang, J., Shin, J., & Xie, S. (2025) · International Conference on Learning Representations · doi:10.48550/arxiv.2410.06940 · Cited by 708
“First, there have been several approaches in generative adversarial network (GAN; Goodfellow et al. 2014) that try to accelerate training with better convergence using pretrained visual encoders (Sauer et al., 2021; Kumari et al., 2022; Sauer et al., 2022; 2023a; Kang et al., 2023).”
Mean Flows for One-step Generative Modeling
Geng, Z., Deng, M., Bai, X., Kolter, J. Z., & He, K. (2025) · neural information processing systems · doi:10.52202/085713-2534 · Cited by 696
Multi-Concept Customization of Text-to-Image Diffusion
Kumari, N., Zhang, B., Zhang, R., Shechtman, E., & Zhu, J.-Y. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.00192 · Cited by 674
“To improve controllability, these models can be conditioned on a class [9,62], image [27,43,73,86], or text prompt [45,67,87].”
MethodLanguage Model Beats Diffusion -- Tokenizer is Key to Visual Generation
Yu, L., Lezama, J., Gundavarapu, N. B., Versari, L., Sohn, K., Minnen, D. C., Cheng, Y., Birodkar, V., Gupta, A., Gu, X., et al. (2024) · International Conference on Learning Representations · doi:10.48550/arxiv.2310.05737 · Cited by 589
“Type Method w/o guidance w/ guidance #Params #Steps FIDÓ ISÒ FIDÓ ISÒ GAN StyleGAN-XL (Sauer et al., 2022) 2.”
MethodElucidating the Design Space of Diffusion-Based Generative Models
Karras, T., Aittala, M., Aila, T., & Laine, S. (2022) · neural information processing systems · doi:10.52202/068431-1926 · Cited by 538
Analyzing and Improving the Training Dynamics of Diffusion Models
Karras, T., Aittala, M., Lehtinen, J., Hellsten, J., Aila, T., & Laine, S. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.02282 · Cited by 530
“Model averaging [27, 49, 57, 68, 72] is an indispensable technique in all high-quality image synthesis methods [2, 11, 23, 29, 31, 33, 46, 48, 54, 60, 63, 65].”
Simple diffusion: End-to-end diffusion for high resolution imagesPreprint
Hoogeboom, E., Heek, J., & Salimans, T. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2301.11093 · Cited by 474
“Other types of generative models have also been successfully applied to image generation (Chang et al., 2022; Sauer et al., 2022; Anonymous, 2023), although modifications such as guidance and low temperature sampling can make it difficult to compare these models fairly.”
Unleashing the Power of Edge-Cloud Generative AI in Mobile Networks: A Survey of AIGC Services
Xu, M., Du, H., Niyato, D., Kang, J., Xiong, Z., Mao, S., Han, Z., Jamalipour, A., Kim, D. I., Shen, X., et al. (2024) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2024.3353265 · Cited by 449
“Image Generation StyleGAN [74], BigGANs [75], StyleGANXL [76], DVD-GAN [77], DALLE [6], DALLE2 [7],”
Scaling up GANs for Text-to-Image Synthesis
Kang, M., Zhu, J.-Y., Zhang, R., Park, J., Shechtman, E., Paris, S., & Park, T. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.00976 · Cited by 405
“GAN2 [42] and 6× larger than StyleGAN-XL [86] and XMC-GAN [103].”
Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation
Sauer, A., Boesel, F., Dockhorn, T., Blattmann, A., Esser, P., & Rombach, R. (2024) · SIGGRAPH Asia Conference Papers · doi:10.1145/3680528.3687625 · Cited by 389
“[Sauer et al. 2023a, 2022].”
Self-citationA Survey on Video Diffusion Models
Xing, Z., Feng, Q., Chen, H., Dai, Q., Hu, H., Xu, H., Wu, Z., & Jiang, Y. (2024) · ACM Computing Surveys · doi:10.1145/3696415 · Cited by 390
“…these endeavors, a series of approaches represented by diffusion models [1–7] have emerged as particularly successful, rapidly supplanting methods based on generative adversarial networks (GANs) [8–12] and auto-regressive Transformers [13–16] to become the predominant approach for image generation.”
Efficient Diffusion Training via Min-SNR Weighting Strategy
Hang, T., Gu, S., Li, C., Bao, J., Chen, D., Hu, H., Geng, X., & Guo, B. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00684 · Cited by 370
StyleGAN-T: Unlocking the Power of GANs for Fast Large-Scale Text-to-Image SynthesisPreprint
Sauer, A., Karras, T., Laine, S., Geiger, A., & Aila, T. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2301.09515 · Cited by 351
“We choose StyleGAN-XL as our baseline architecture because of its strong performance in class-conditional Ima-geNet synthesis (Sauer et al., 2022).”
Self-citationReconstructing the Mind's Eye: fMRI-to-Image with Contrastive Learning and Diffusion Priors
Scotti, P. S., Banerjee, A., Goode, J., Shabalin, S., Nguyen, A., Cohen, E., Dempster, A. J., Verlinde, N., Yundler, E., Weisberg, D., et al. (2023) · neural information processing systems · doi:10.52202/075280-1073 · Cited by 339
FLAME: Free-Form Language-Based Motion Synthesis & Editing
Kim, J.-H., Kim, J., & Choi, S. (2023) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v37i7.25996 · Cited by 316
The Role of ImageNet Classes in Fréchet Inception DistancePreprint
Kynkäänniemi, T., Karras, T., Aittala, M., Aila, T., & Lehtinen, J. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2203.06026 · Cited by 311
“Viable alternative feature spaces include CLIP (Radford et al., 2021; Sauer et al., 2021), self-supervised SwAV (Caron et al., 2020; Morozov et al., 2021), and an uninitialized network (Naeem et al., 2020; Sauer et al., 2022).”
MethodDataset Distillation: A Comprehensive Review
Yu, R., Liu, S., & Wang, X. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3323376 · Cited by 308
“…V-A, we expect future works to design specific DD methods for more tasks in computer vision, e.g., semantic segmentation [172], [173], objective detection [174], and generative modeling [175], [176], [177], natural language processing, e.g., machine translation [178], and multi-modal scenarios [6].”
BackgroundExposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
Stein, G., Cresswell, J. C., Hosseinzadeh, R., Sui, Y., Ross, B. L., Villecroze, V., Liu, Z., Caterini, A. L., Taylor, E., & Loaiza-Ganem, G. (2023) · neural information processing systems · doi:10.52202/075280-0165 · Cited by 294
Improved Techniques for Training Consistency Models
Song, Y., & Dhariwal, P. (2024) · International Conference on Learning Representations · doi:10.48550/arxiv.2310.14189 · Cited by 283
“Here we exclude methods based on FastGAN (Liu et al., 2020; Sauer et al., 2021) or StyleGAN-XL (Sauer et al., 2022) from our consideration, because both utilize ImageNet pre-trained feature extractors in their discriminators.”
MethodAll are Worth Words: A ViT Backbone for Diffusion Models
Bao, F., Nie, S., Xue, K., Cao, Y., Li, C., Su, H., & Zhu, J. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.02171 · Cited by 274
On the Importance of Noise Scheduling for Diffusion ModelsPreprint
Chen, T. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2301.10972 · Cited by 274
“We note that state-of-the-art GANs [17] can achieve similar or better performance but with multi-stage generation, as well as classifier-guidance [3], which we do not use for quantitative evaluation.”
Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion
Kim, D., Lai, C.-H., Liao, W., Murata, N., Takida, Y., Uesaka, T., He, Y., Mitsufuji, Y., & Ermon, S. (2024) · International Conference on Learning Representations · doi:10.48550/arxiv.2310.02279 · Cited by 263
“The output of the cross-scale mixing is a feature pyramid consisting of four feature maps at different resolutions (Sauer et al., 2022).”
MethodThe Stable Signature: Rooting Watermarks in Latent Diffusion Models
Fernandez, P., Couairon, G., Jégou, H., Douze, M., & Furon, T. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.02053 · Cited by 251
“Image generation has long been dominated by GANs, still state-of-the-art on many datasets [37, 38, 39, 76, 87].”
BackgroundAdversarial Diffusion Distillation
Sauer, A., Lorenz, D., Blattmann, A., & Rombach, R. (2025) · Lecture Notes in Computer Science · doi:10.1007/978-3-031-73016-0_6 · Cited by 242
Self-citationImage Denoising: The Deep Learning Revolution and Beyond—A Survey Paper
Elad, M., Kawar, B., & Vaksman, G. (2023) · SIAM Journal on Imaging Sciences · doi:10.1137/23m1545859 · Cited by 245
SEINE: Short-to-Long Video Diffusion Model for Generative Transition and Prediction
Chen, X., Wang, Y., Zhang, L., Zhuang, S., Ma, X., Yu, J., Wang, Y., Lin, D., Qiao, Y., & Liu, Z. (2023) · International Conference on Learning Representations · doi:10.48550/arxiv.2310.20700 · Cited by 227
“This approach has been widely explored in various works, e.g., style transfer (Chen et al., 2018), object transfiguration (Sauer et al., 2022; Kang et al., 2023).”
BackgroundUFOGen: You Forward Once Large Scale Text-to-Image Generation via Diffusion GANs
Xu, Y., Zhao, Y., Xiao, Z., & Hou, T. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.00783 · Cited by 225
“Later, with the evolution of more sophisticated GAN architectures [24, 25, 53], GANs trained on large datasets have shown promising results in the domain of text-to-image generation [22, 54, 72].”
BackgroundFlow Matching in Latent SpacePreprint
Dao, Q. T., Phung, H., Nguyen, B., & Tran, A. T. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2307.08698 · Cited by 223
HPSv3: Towards Wide-Spectrum Human Preference Score
Ma, Y., Wu, X., Sun, K., & Li, H. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.01400 · Cited by 227
StyleGAN-Human: A Data-Centric Odyssey of Human Generation
Fu, J., Li, S., Jiang, Y., Lin, K.-Y., Qian, C., Loy, C. C., Wu, W., & Liu, Z. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19787-1_1 · Cited by 213
Simplifying, Stabilizing and Scaling Continuous-Time Consistency Models
Lu, C. C., & Song, Y. (2025) · International Conference on Learning Representations · doi:10.48550/arxiv.2410.11081 · Cited by 211
“Notably, the 1-step FID of sCD-XXL on ImageNet 512 × 512 surpasses that of StyleGAN-XL (Sauer et al., 2022) and VAR (Tian et al., 2024a).”
HyperDiffusion: Generating Implicit Neural Fields with Weight-Space Diffusion
Erkoç, Z., Ma, F., Shan, Q., Nießner, M., & Dai, A. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.01315 · Cited by 194
“Generative Adversarial Networks (GANs) Generative Adversarial Networks (GANs) [18] has been shown to be capable of generating high-resolution 2D images, most notably StyleGAN [24, 25, 23, 42].”
BackgroundGeneralizing Dataset Distillation via Deep Generative Prior
Cazenavette, G., Wang, T., Torralba, A. B., Efros, A. A., & Zhu, J.-Y. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.00364 · Cited by 190
“Given the depth of StyleGAN-XL [51], there are many possible latent spaces that GLaD can use to parameterize our synthetic dataset.”
BackgroundMethodA review on Generative Adversarial Networks for image generation
de Souza, V. L. T., Marques, B. A. D., Batagelo, H. C., & Gois, J. P. (2023) · Computers & Graphics · doi:10.1016/j.cag.2023.05.010 · Cited by 183
VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids
Schwarz, K., Sauer, A., Niemeyer, M., Liao, Y., & Geiger, A. (2022) · neural information processing systems · doi:10.52202/068431-2464 · Cited by 172
Self-citationDiffiT: Diffusion Vision Transformers for Image Generation
Hatamizadeh, A., Song, J., Liu, G., Kautz, J., & Vahdat, A. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73242-3_3 · Cited by 170
Generative Modeling via DriftingPreprint
Deng, M., Li, H., Li, T., Du, Y., & He, K. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2602.04770 · Cited by 169
“Comparing with other one-step, pixel-space methods (GANs), our method achieves 1.61 FID using only 87G FLOPs; by comparison, StyleGAN-XL produces 2.30 FID using 1574G FLOPs.”
InfluentialOne-Step Diffusion with Distribution Matching Distillation
Yin, T., Gharbi, M., Zhang, R., Shechtman, E., Durand, F., Freeman, W. T., & Park, T. (2024) · 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52733.2024.00632 · Cited by 163
Fast Training of Diffusion Models with Masked Transformers
Zheng, H., Nie, W., Vahdat, A., & Anandkumar, A. (2024) · Transactions on machine learning research · doi:10.48550/arxiv.2306.09305 · Cited by 161
Xie, E., Yao, L., Shi, H., Liu, Z., Zhou, D., Liu, Z., Li, J., & Li, Z. (2023) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.00390 · Cited by 157
The GAN is dead; long live the GAN! A Modern GAN Baseline
Huang, Y., Gokaslan, A., Kuleshov, V., & Tompkin, J. (2024) · neural information processing systems · doi:10.52202/079017-1402 · Cited by 160
Gao, P., Zhuo, L., Liu, D., Du, R., Xu, L., Qiu, L., Zhang, Y., Lin, C., Huang, R., Geng, S., et al. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2405.05945 · Cited by 159
Song, Y., Dhariwal, P., Chen, M., & Sutskever, I. (2023) · arXiv · doi:10.48550/arxiv.2303.01469 · Cited by 155
A GAN-Based Data Augmentation Method for Imbalanced Multi-Class Skin Lesion Classification
Su, Q., Hamed, H. N. A., Isa, M. A., Xue, H., & Dai, X. (2024) · IEEE Access · doi:10.1109/access.2024.3360215 · Cited by 156
“This study use improved Precision and Recall [73], as this study aim to mitigate the the mode collapse problem of minority classes and low recall indicates mode collapse [74].”
MethodAffordance Diffusion: Synthesizing Hand-Object Interactions
Ye, Y., Li, X., Gupta, A., Mellon, S. D., Birchfield, S. T., Song, J., Tulsiani, S., & Liu, S. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.02153 · Cited by 154
“On the one hand, there has been a lot of progress in image generation, such as synthesizing realistic high-resolution images spanning a wide range of object categories [44, 74] from human faces to ImageNet classes.”
BackgroundSimDA: Simple Diffusion Adapter for Efficient Video Generation
Xing, Z., Dai, Q., Hu, H., Wu, Z., & Jiang, Y. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.00748 · Cited by 148
MDTv2: Masked Diffusion Transformer is a Strong Image SynthesizerPreprint
Gao, S., Zhou, P., Cheng, M.-M., & Yan, S. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2303.14389 · Cited by 146
Visual Prompt Tuning for Generative Transfer Learning
Sohn, K., Chang, H., Lezama, J., Polanía, L. F., Zhang, H., Hao, Y., Essa, I. A., & Jiang, L. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.01900 · Cited by 145
RenAIssance: A Survey Into AI Text-to-Image Generation in the Era of Large Model
Bie, F., Yang, Y., Zhou, Z., Ghanem, A., Zhang, M., Yao, Z., Wu, X., Holmes, C., Golnari, P. A., Clifton, D. A., et al. (2024) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2024.3522305 · Cited by 141
OmniTokenizer: A Joint Image-Video Tokenizer for Visual Generation
Wang, J., Jiang, Y., Yuan, Z., Peng, B., Wu, Z., & Jiang, Y.-G. (2024) · neural information processing systems · doi:10.52202/079017-0887 · Cited by 138
State‐of‐the‐Art in the Architecture, Methods and Applications of StyleGAN
Bermano, A. H., Gal, R., Alaluf, Y., Mokady, R., Nitzan, Y., Tov, O., Patashnik, O., & Cohen‐Or, D. (2022) · Computer Graphics Forum · doi:10.1111/cgf.14503 · Cited by 141
Generative Modelling With Inverse Heat DissipationPreprint
Rissanen, S., Heinonen, M., & Solin, A. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2206.13397 · Cited by 132
“Other famous examples are the progressive GAN [34] and later the StyleGAN architectures [35–37 , 68 Multiscale ideas in the context of autoregressive models have been proposed in the past, starting with the multi-scale pixelRNN model [78].”
BackgroundScalable High-Resolution Pixel-Space Image Synthesis with Hourglass Diffusion TransformersPreprint
Crowson, K., Baumann, S. A., Birch, A., Abraham, T. M., Kaplan, D. Z., & Shippole, E. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2401.11605 · Cited by 130
“We find that our model is competitive regarding FID with high-resolution transformer GANs such as HiT (Zhao et al., 2021) or StyleSwin (Zhang et al., 2022a), but does not reach the same FID as state-of-the-art GANs such as StyleGAN-XL (Sauer et al., 2022).”
InfluentialEfficient Diffusion Models for Vision: A SurveyPreprint
Ulhaq, A., & Akhtar, N. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2210.09292 · Cited by 127
Geng, Z., Pokle, A., Luo, W., Lin, J., & Kolter, J. Z. (2024) · International Conference on Learning Representations · doi:10.48550/arxiv.2406.14548 · Cited by 127
“BigGAN [5] 326.66 1 StyleGAN2-ADA [25] 305.92 1 StyleGAN-XL [59] 204.”
InfluentialFlow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow
Liu, X., Gong, C., & Liu, Q. (2022) · arXiv · doi:10.48550/arxiv.2209.03003 · Cited by 124
PixelFlow: Pixel-Space Generative Models with FlowPreprint
Chen, S., Ge, C., Zhang, S., Sun, P., & Luo, P. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2504.07963 · Cited by 127
Simpler Diffusion: 1.5 FID on ImageNet512 with pixel-space diffusion
Hoogeboom, E., Mensink, T., Heek, J., Lamerigts, K., Gao, R., & Salimans, T. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.01683 · Cited by 120
NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion Models
Kim, S. W., Brown, B., Yin, K., Kreis, K., Schwarz, K., Li, D., Rombach, R., Torralba, A. B., & Fidler, S. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.00821 · Cited by 119
“2D Generative Models In past years, generative adversarial networks (GANs) [3, 17, 27, 41, 58] and likelihoodbased approaches [33, 49, 51, 67] enabled high-resolution photorealistic image synthesis.”
Self-citationData synthesis and adversarial networks: A review and meta-analysis in cancer imaging
Osuala, R. O., Kushibar, K., Garrucho, L., Linardos, A., Szafranowska, Z., Klein, S., Glocker, B., Díaz, O., & Lekadir, K. (2022) · Medical Image Analysis · doi:10.1016/j.media.2022.102704 · Cited by 131
Distilling Diffusion Models Into Conditional GANs
Kang, M., Zhang, R., Barnes, C., Paris, S., Kwak, S., Park, J., Shechtman, E., Zhu, J.-Y., & Park, T. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73390-1_25 · Cited by 116
CADS: Unleashing the Diversity of Diffusion Models through Condition-Annealed Sampling
Sadat, S., Buhmann, J., Bradley, D., Hilliges, O., & Weber, R. M. (2023) · International Conference on Learning Representations · doi:10.48550/arxiv.2310.17347 · Cited by 115
Deepfake Media Forensics: Status and Future Challenges
Amerini, I., Barni, M., Battiato, S., Bestagini, P., Boato, G., Bruni, V., Caldelli, R., De Natale, F., De Nicola, R., Guarnera, L., et al. (2025) · Journal of Imaging · doi:10.3390/jimaging11030073 · Cited by 113
“StyleGAN-XL [36] is an extension of the StyleGAN family, designed to generate high-quality images on diverse and unstructured datasets, particularly ImageNet.”
Diffusion Models Without Attention
Yan, J. N., Gu, J., & Rush, A. M. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.00787 · Cited by 113
“We believe removing the attention bottleneck should open up the possibility of applications in other areas that requires long-range diffusion, for example high-fidelity audio, video, or 3D modeling.”
The ArtBench Dataset: Benchmarking Generative Models with ArtworksPreprint
Liao, P., Li, X., Liu, X., & Keutzer, K. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2206.11404 · Cited by 111
“Deep generative models [3, 9, 15, 20, 41, 46] are capable of modeling complex, high dimensional distributions, and one of the prominent applications is the synthesizing diverse and high-fidelity images.”
Align Your Flow: Scaling Continuous-Time Flow Map Distillation
Sabour, A., Fidler, S., & Kreis, K. (2025) · neural information processing systems · doi:10.52202/085713-4900 · Cited by 109
Understanding GANs: fundamentals, variants, training challenges, applications, and open problems
Ahmad, Z., Jaffri, Z. U. A., Chen, M., & Bao, S. (2024) · Multimedia Tools and Applications · doi:10.1007/s11042-024-19361-y · Cited by 110
DiffMorpher: Unleashing the Capability of Diffusion Models for Image Morphing
Zhang, K., Zhou, Y., Xu, X., Dai, B., & Pan, X. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.00756 · Cited by 108
Improved Mean Flows: On the Challenges of Fastforward Generative ModelsPreprint
Geng, Z., Lu, Y., Wu, Z., Shechtman, E., Kolter, J. Z., & He, K. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2512.02012 · Cited by 108
Face Generation and Editing With StyleGAN: A Survey
Melnik, A., Miasayedzenkau, M., Makaravets, D., Pirshtuk, D., Akbulut, E., Holzmann, D., Renusch, T., Reichert, G., & Ritter, H. (2024) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2024.3350004 · Cited by 105
“neural network adapters for integration with other architectures [118] [119] [120], scaling StyleGAN to large diverse datasets [121] or new interactive image”
BackgroundStudioGAN: A Taxonomy and Benchmark of GANs for Image Synthesis
Kang, M., Shin, J., & Park, J. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3306436 · Cited by 104
GIVT: Generative Infinite-Vocabulary Transformers
Tschannen, M., Eastwood, C., & Mentzer, F. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72998-0_17 · Cited by 104
DDT: Decoupled Diffusion TransformerPreprint
Wang, S., Tian, Z., Huang, W., & Wang, L. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2504.05741 · Cited by 104
“Diffusion models [19, 21, 29, 30, 41] have recently emerged as highly promising solutions to learn the underlying data distribution in image generation, outperforming the GAN-based models [3, 40] and Auto-Regressive models [5, 43, 51].”
A Systematic Survey of Regularization and Normalization in GANs
Li, Z., Usman, M., Tao, R., Xia, P., Wang, C., Chen, H., & Li, B. (2023) · ACM Computing Surveys · doi:10.1145/3569928 · Cited by 101
“StyleGAN-XL (2022 [110]) Conditional Generation None Translation Cutout None SN:D IN”
BackgroundAutoregressive Models in Vision: A Survey
Xiong, J., Liu, G., Huang, L., Wu, C., Wu, T., Mu, Y., Yao, Y., Shen, H., Wan, Z., Huang, J., et al. (2025) · Transactions on machine learning research · doi:10.48550/arxiv.2411.05902 · Cited by 98
RandAR: Decoder-only Autoregressive Visual Generation in Random Orders
Pang, Z., Zhang, T., Luan, F., Man, Y., Tan, H., Zhang, K., Freeman, W. T., & Wang, Y.-X. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.00014 · Cited by 97
Truncated Diffusion Probabilistic Models and Diffusion-based Adversarial Auto-EncodersPreprint
Zheng, H., He, P., Chen, W., & Zhou, M. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2202.09671 · Cited by 98
“Here we adopt the TDPM+ configuration, where we use a pre-trained ADM (Dhariwal & Nichol, 2021) checkpoint for t < T Trunc and train a StyleGAN-XL (Sauer et al., 2022) based implicit model at t = T Trunc (for simplicity, we choose to not use the progressive growing pipeline of StyleGAN-XL; See…”
MethodMicrostructure reconstruction using diffusion-based generative models
Lee, K., & Yun, G. J. (2023) · Mechanics of Advanced Materials and Structures · doi:10.1080/15376494.2023.2198528 · Cited by 95
“In addition, StyleGAN [37,38] has the lowest Fr (cid:1) echet inception distance (FID) scores with ImageNet 64 (cid:1) 64 dataset as shown in Table 1.”
Semi-Parametric Neural Image SynthesisPreprint
Blattmann, A., Rombach, R., Oktay, K., Müller, J., & Ommer, B. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2204.11824 · Cited by 95
“Using these parameters for RDM-IN results in a model which even achieves similar FID scores than state of the class-conditional models on ImageNet [63, 15, 70] without requiring any labels during training or inference.”
BackgroundMethodInfluentialFiT: Flexible Vision Transformer for Diffusion ModelPreprint
Lu, Z., Wang, Z., Huang, D., Wu, C., Liu, X., Ouyang, W., & Bai, L. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2402.12376 · Cited by 94
“…from the FiT in Fig 1, and we compare against some state-of-the-art class-conditional generative models: BigGAN (Brock et al., 2018), StyleGAN-XL (Sauer et al., 2022), MaskGIT (Chang et al., 2022), CDM (Ho et al., 2022), U-ViT (Bao et al., 2023), ADM (Dhariwal & Nichol, 2021), LDM (Rombach et…”
MethodUnifying Diffusion Models' Latent Space, with Applications to CycleDiffusion and GuidancePreprint
Wu, C. H., & la Torre, F. D. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2210.05559 · Cited by 92
Masked Diffusion Transformer is a Strong Image Synthesizer
Gao, S., Zhou, P., Cheng, M.-M., & Yan, S. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.02117 · Cited by 91
ControlNet-XS: Rethinking the Control of Text-to-Image Diffusion Models as Feedback-Control Systems
Zavadski, D., Feiden, J.-F., & Rother, C. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73223-2_20 · Cited by 88
“Generative Adversarial Networks (GANs) [20] are probably the most established generative models for unconditional [34–36], class-conditional [72] and text-conditional [32,63,64,71,78,83,87,95] image generation as well as domain adaptation [49,82,94].”
BackgroundApplying Deep Learning Methods for Mammography Analysis and Breast Cancer Detection
Prodan, M., Paraschiv, E. A., & Stanciu, A. (2023) · Applied Sciences · doi:10.3390/app13074272 · Cited by 87
“Therefore, 1000 synthetic images were generated as positive examples using the StyleGAN-XL model [36] in the third stage.”
Community Forensics: Using Thousands of Generators to Train Fake Image Detectors
Park, J., & Owens, A. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.00772 · Cited by 91
One-Step Diffusion Distillation through Score Implicit Matching
Luo, W., Huang, Z., Geng, Z., Kolter, J. Z., & Qi, G.-J. (2024) · neural information processing systems · doi:10.52202/079017-3664 · Cited by 87
Scaling Diffusion Transformers to 16 Billion ParametersPreprint
Fei, Z., Fan, M., Yu, C., Li, D., & Huang, J. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2407.11633 · Cited by 89
Latent Diffusion Model without Variational Autoencoder
Shi, M., Wang, H., Zheng, W., Yuan, Z., Wu, X., Wang, X., Wan, P., Zhou, J., & Lu, J. (2025) · International Conference on Learning Representations · doi:10.48550/arxiv.2510.15301 · Cited by 86
“…et al., 2014; Radford et al., 2016; Arjovsky et al., 2017; Gulrajani et al., 2017; Karras et al., 2019; Zhu et al., 2017; Karras et al., 2018; Sauer et al., 2022) generate realistic images via adversarial training but often suffer from mode collapse, instability, and poor interpretability.”
PixNerd: Pixel Neural Field Diffusion
Wang, S., Gao, Z., Zhu, C., Huang, W., & Wang, L. (2025) · International Conference on Learning Representations · doi:10.48550/arxiv.2507.23268 · Cited by 85
“VAEs typically requires adversarial training[4, 5, 1, 2, 3] and additional supervision [6], introducing complex optimization challenges.”
One-step Latent-free Image Generation with Pixel Mean FlowsPreprint
Lu, Y., Lu, S., Sun, Q., Zhao, H., Jiang, Z., Wang, X., Li, T., Geng, Z., & He, K. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2601.22158 · Cited by 84
“For example, StyleGAN-XL (Sauer et al., 2022) costs 1574 Gflops per forward, 5.8 × more than our pMF-H/16.”
InfluentialParallelized Autoregressive Visual Generation
Wang, Y., Ren, S., Lin, Z., Han, Y., Guo, H., Yang, Z., Zou, D., Feng, J., & Liu, X. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.01209 · Cited by 83
“2 presents comprehensive comparisons of class-conditional image generation with various state-of-the-art methods, including GAN [3, 20, 44](one-shot generation), Diffusion [11, 17, 37, 42] (iterative denoising), Mask [5] (mask token prediction), VAR [53] (next-scale prediction), MAR [27]…”
What matters for Representation Alignment: Global Information or Spatial Structure?
Singh, J., Leng, X., Wu, Z., Zheng, L., Zhang, R., Shechtman, E., & Xie, S. Q. (2025) · International Conference on Learning Representations · doi:10.48550/arxiv.2512.10794 · Cited by 83
“In GANs (Goodfellow et al., 2020), pretrained features have not only improved convergence speed but also enabled scaling to large datasets, as demonstrated by StyleGAN-XL (Sauer et al., 2022) and StyleGAN-T (Sauer et al., 2023) with CLIP (Radford et al., 2021) features.”
Beyond Next-Token: Next-X Prediction for Autoregressive Visual Generation
Ren, S., Yu, Q., Ju, H., Shen, X., Yuille, A., & Chen, L.-C. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.01464 · Cited by 84
“Beyond autoregressive modeling, diffusion [19, 45, 52, 53] and flow matching [12, 16, 28, 30, 40, 41, 51, 64] have surpassed Generative Adversarial Networks (GANs) [14, 50] by employing multi-step denoising.”
Improved Masked Image Generation with Token-Critic
Lezama, J., Chang, H., Jiang, L., & Essa, I. A. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20050-2_5 · Cited by 80
Metadata
- Authors
- Sauer, A., Schwarz, K., & Geiger, A.
- Affiliations
- Max Planck Institute for Intelligent Systems (DE)
- Venue
- Special Interest Group on Computer Graphics and Interactive Techniques Conference Proceedings
- Publisher
- ACM
- Published
- 2022-08-07
- Pages
- 1-10
- OpenAlex
- W4285981784
- Cited by
- 353
- References
- 19
- Citations by year
- 2026: 42 · 2025: 84 · 2024: 122 · 2023: 86 · 2022: 19
- FWCI
- 28.20
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- ACM SIGGRAPH Conference (SIGGRAPH) (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 Graphics and Computer-Aided Design › Computer Graphics and Visualization Techniques
- Physical Sciences › Computer Science › Computer Graphics and Computer-Aided Design › 3D Shape Modeling and Analysis
- Keywords
- StyleGAN3, StyleGAN, ImageNet, image inversion, image editing, high-resolution image generation, generative adversarial networks
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.