Masked Diffusion Transformer is a Strong Image Synthesizer
Open accessAbstract
Despite its success in image synthesis, we observe that diffusion probabilistic models (DPMs) often lack contextual reasoning ability to learn the relations among object parts in an image, leading to a slow learning process. To solve this issue, we propose a Masked Diffusion Transformer (MDT) that introduces a mask latent modeling scheme to explicitly enhance the DPMs’ ability to contextual relation learning among object semantic parts in an image. During training, MDT operates in the latent space to mask certain tokens. Then, an asymmetric masking diffusion transformer is designed to predict masked tokens from unmasked ones while maintaining the diffusion generation process. Our MDT can reconstruct the full information of an image from its incomplete contextual input, thus enabling it to learn the associated relations among image tokens. Experimental results show that MDT achieves superior image synthesis performance, e.g., a new SOTA FID score in the ImageNet data set, and has about 3× faster learning speed than the previous SOTA DiT. The source code is released at https://github.com/sail-sg/MDT.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
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?
Recent research indicates that token reduction has the potential to significantly improve training stability when integrated into the pre-training phase, suggesting that selective token utilization during training can lead to more robust model learning.
“While token reduction has traditionally been employed as a post-training optimization to enhance inference efficiency, recent research indicates its potential to significantly improve training stability when integrated into the pre-training phase [36, 83, 86], suggesting that selective token utilization during training can lead to more robust model learning, as shown in Fig. 2d.”
From Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodality (2025) · cites this paper as [36]
Cites
U-Net: Convolutional Networks for Biomedical Image Segmentation
Ronneberger, O., Fischer, P., & Brox, T. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24574-4_28 · Cited by 99,404 · see citing works
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 48,293 · see citing works
Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., & Guo, B. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00986 · Cited by 35,534 · see citing works
High-Resolution Image Synthesis with Latent Diffusion Models
Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.01042 · Cited by 15,880 · see citing works
Masked Autoencoders Are Scalable Vision Learners
He, K., Chen, X., Xie, S., Li, Y., Dollar, P., & Girshick, R. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.01553 · Cited by 8,489 · see citing works
GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., & Hochreiter, S. (2017) · arXiv · doi:10.48550/arxiv.1706.08500 · Cited by 8,365 · see citing works
Large Scale GAN Training for High Fidelity Natural Image Synthesis
Brock, A., Donahue, J., & Simonyan, K. (2018) · arXiv · doi:10.48550/arxiv.1809.11096 · Cited by 3,717 · see citing works
Diffusion Models Beat GANs on Image Synthesis
Dhariwal, P., & Nichol, A. (2021) · arXiv · doi:10.48550/arxiv.2105.05233 · Cited by 3,362 · see citing works
Attention mechanisms in computer vision: A survey
Guo, M.-H., Xu, T.-X., Liu, J.-J., Liu, Z.-N., Jiang, P.-T., Mu, T.-J., Zhang, S.-H., Martin, R. R., Cheng, M.-M., & Hu, S.-M. (2022) · Computational Visual Media · doi:10.1007/s41095-022-0271-y · Cited by 2,566 · cited 1× · see citing works
“Several design improvements, e.g ., adding more attention heads [18], BigGAN [4] residual block, and adaptive group normalization, are proposed in [11] to further enhance the generation ability of the UNet.”
MethodSelf-citationScalable 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 · see citing works
Score-Based Generative Modeling through Stochastic Differential Equations
Song, Y., Sohl-Dickstein, J., Kingma, D. P., Kumar, A., Ermon, S., & Poole, B. (2020) · arXiv · doi:10.48550/arxiv.2011.13456 · Cited by 1,953 · see citing works
High-Resolution Image Synthesis with Latent Diffusion Models
Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2021) · arXiv · doi:10.48550/arxiv.2112.10752 · Cited by 737 · see citing works
MaskGIT: Masked Generative Image Transformer
Chang, H., Zhang, H., Jiang, L., Liu, C., & Freeman, W. T. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52688.2022.01103 · Cited by 446 · see citing works
Progressive Distillation for Fast Sampling of Diffusion Models
Salimans, T., & Ho, J. (2022) · arXiv · doi:10.48550/arxiv.2202.00512 · Cited by 287 · see citing works
Large-Scale Unsupervised Semantic Segmentation
Gao, S., Li, Z.-Y., Yang, M.-H., Cheng, M.-M., Han, J., & Torr, P. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3218275 · Cited by 88 · see citing works
Self-citation
Flagged first, then by citations
Cited by
Sorted by citations
Switch Diffusion Transformer: Synergizing Denoising Tasks with Sparse Mixture-of-Experts
Park, B., Go, H., Kim, J.-Y., Woo, S., Ham, S., & Kim, C. (2025) · Lecture Notes in Computer Science · doi:10.1007/978-3-031-73668-1_27 · Cited by 9 · see citing works
Kong, Z., Li, Y., Zeng, F., Xin, L., Messica, S., Lin, X., Zhao, P., Kellis, M., Tang, H., & Zitnik, M. (2025) · arXiv · doi:10.48550/arxiv.2505.18227
“It can be applied to both diffusion models [14, 97, 136] and diffusion transformers [36, 189].”
- Authors
- Gao, S., Zhou, P., Cheng, M.-M., & Yan, S.
- Affiliations
- Nankai University (CN)
- Sea AI Lab
- Venue
- 2023 IEEE/CVF International Conference on Computer Vision (ICCV)
- Publisher
- IEEE
- Published
- 2023-10-01
- Pages
- 23107-23116
- OpenAlex
- W4390873332
- Cited by
- 91
- References
- 15
- Citations by year
- 2026: 28 · 2025: 37 · 2024: 22 · 2023: 4
- FWCI
- 9.74
- Type
- conference paper
- Language
- en
- Open access
- Locations
- IEEE/CVF International Conference on Computer Vision (ICCV) (publishedVersion)
- Singapore Management University Institutional Knowledge (InK) (Singapore Management University) (submittedVersion, cc-by-nc-nd)
- Funders
- Fundamental Research Funds for the Central Universities
- 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 › Multimodal Machine Learning Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Keywords
- diffusion probabilistic models, image generation, image tokens, learning speed, DiT
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
