mlsanityData to 7 Oct 2026

Improving Sample Quality of Diffusion Models Using Self-Attention Guidance

arXiv 2210.00939 ↗First posted 3 Oct 2022cs.CV cs.AI cs.LGcited by 151

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Denoising diffusion models (DDMs) have attracted attention for their exceptional generation quality and diversity. This success is largely attributed to the use of class- or text-conditional diffusion guidance methods, such as classifier and classifier-free guidance. In this paper, we present a more comprehensive perspective that goes beyond the traditional guidance methods. From this generalized perspective, we introduce novel condition- and training-free strategies to enhance the quality of generated images. As a simple solution, blur guidance improves the suitability of intermediate samples for their fine-scale information and structures, enabling diffusion models to generate higher quality samples with a moderate guidance scale. Improving upon this, Self-Attention Guidance (SAG) uses the intermediate self-attention maps of diffusion models to enhance their stability and efficacy. Specifically, SAG adversarially blurs only the regions that diffusion models attend to at each iteration and guides them accordingly. Our experimental results show that our SAG improves the performance of various diffusion models, including ADM, IDDPM, Stable Diffusion, and DiT. Moreover, combining SAG with conventional guidance methods leads to further improvement.

Key facts. Cited by 151 papers in the map (data through 7 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 41 of its 58 references that are in the map. First posted on arXiv 3 Oct 2022.

Builds on

41 of 58 references.

Closest ancestors · 12

Same direction, newest first

  1. 2022Blurring Diffusion Models12 Sep
  2. 2022Classifier-Free Diffusion Guidance26 Jul
  3. 2022Progressive Deblurring of Diffusion Models for Coarse-to-Fine Image Synthesis16 Jul
  4. 2022Improved Vector Quantized Diffusion Models31 May
  5. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  6. 2021Vector Quantized Diffusion Model for Text-to-Image Synthesis29 Nov
  7. 2021Cascaded Diffusion Models for High Fidelity Image Generation30 May
  8. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  9. 2021Improved Denoising Diffusion Probabilistic Models18 Feb
  10. 2020Score-Based Generative Modeling through Stochastic Differential Equations26 Nov
  11. 2020Denoising Diffusion Implicit Models6 Oct
  12. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 29

  1. 2017Attention Is All You Needcited by 59,932
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  4. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  5. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  6. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  7. 2017Mask R-CNNcited by 7,259
  8. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  9. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  10. 2022Hierarchical Text-Conditional Image Generation with CLIP Latentscited by 6,539
  11. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,322
  12. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
17 more
  1. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 3,595
  2. 2020Taming Transformers for High-Resolution Image Synthesiscited by 3,229
  3. 2016Improved Techniques for Training GANscited by 2,938
  4. 2022Prompt-to-Prompt Image Editing with Cross Attention Controlcited by 2,409
  5. 2015LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loopcited by 1,350
  6. 2019Improved Precision and Recall Metric for Assessing Generative Modelscited by 925
  7. 2018Self-Attention Generative Adversarial Networkscited by 809
  8. 2021Label-Efficient Semantic Segmentation with Diffusion Modelscited by 495
  9. 2021Generating Images with Sparse Representationscited by 262
  10. 2022Generative Modelling With Inverse Heat Dissipationcited by 164
  11. 2021TransGAN: Two Pure Transformers Can Make One Strong GAN, and That Can Scale Upcited by 162
  12. 2022Pretraining is All You Need for Image-to-Image Translationcited by 160
  13. 2021StyleSwin: Transformer-based GAN for High-resolution Image Generationcited by 152
  14. 2022Discovering Object Masks with Transformers for Unsupervised Semantic Segmentationcited by 54
  15. 2022Unsupervised Semantic Segmentation with Self-supervised Object-centric Representationscited by 54
  16. 2020SSD-GAN: Measuring the Realness in the Spatial and Spectral Domainscited by 38
  17. 2022WaveGAN: Frequency-aware GAN for High-Fidelity Few-shot Image Generationcited by 22

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 23; 2024: 46; 2025: 49; 2026: 32.

151 papers.

Sort:

By area:

  1. 2023TokenFlow: Consistent Diffusion Features for Consistent Video Editingcited by 440
  2. 2024Guiding a Diffusion Model with a Bad Version of Itselfcited by 286
  3. 2024StoryDiffusion: Consistent Self-Attention for Long-Range Image and Video Generationcited by 240
  4. 2023Let 2D Diffusion Model Know 3D-Consistency for Robust Text-to-3D Generationcited by 133
  5. 2023Editing Implicit Assumptions in Text-to-Image Diffusion Modelscited by 128
  6. 2023CADS: Unleashing the Diversity of Diffusion Models through Condition-Annealed Samplingcited by 124
  7. 2023Space-Time Diffusion Features for Zero-Shot Text-Driven Motion Transfercited by 124
  8. 2023RAVE: Randomized Noise Shuffling for Fast and Consistent Video Editing with Diffusion Modelscited by 113
  9. 2024Golden Noise for Diffusion Models: A Learning Frameworkcited by 91
  10. 2023Dynamic Prompt Learning: Addressing Cross-Attention Leakage for Text-Based Image Editingcited by 76
  11. 2024Attentive Eraser: Unleashing Diffusion Model's Object Removal Potential via Self-Attention Redirection Guidancecited by 63
  12. 2023On the Design Fundamentals of Diffusion Models: A Surveycited by 59
38 more
  1. 2023DreamSync: Aligning Text-to-Image Generation with Image Understanding Feedbackcited by 57
  2. 2023Enhancing Deep Reinforcement Learning: A Tutorial on Generative Diffusion Models in Network Optimizationcited by 56
  3. 2024Smoothed Energy Guidance: Guiding Diffusion Models with Reduced Energy Curvature of Attentioncited by 55
  4. 2024Diffusion Models and Representation Learning: A Surveycited by 53
  5. 2024No Training, No Problem: Rethinking Classifier-Free Guidance for Diffusion Modelscited by 52
  6. 2023DirecT2V: Large Language Models are Frame-Level Directors for Zero-Shot Text-to-Video Generationcited by 47
  7. 2024SwiftBrush v2: Make Your One-step Diffusion Model Better Than Its Teachercited by 45
  8. 2024Zigzag Diffusion Sampling: Diffusion Models Can Self-Improve via Self-Reflectioncited by 38
  9. 2025Stochastic Self-Guidance for Training-Free Enhancement of Diffusion Modelscited by 35
  10. 2024Alignment of Diffusion Models: Fundamentals, Challenges, and Futurecited by 34
  11. 2024Control Color: Multimodal Diffusion-based Interactive Image Colorizationcited by 33
  12. 2024Spatiotemporal Skip Guidance for Enhanced Video Diffusion Samplingcited by 33
  13. 2023Debiasing Scores and Prompts of 2D Diffusion for View-consistent Text-to-3D Generationcited by 32
  14. 2023SUR-adapter: Enhancing Text-to-Image Pre-trained Diffusion Models with Large Language Modelscited by 29
  15. 2023Brain Diffusion for Visual Exploration: Cortical Discovery using Large Scale Generative Modelscited by 29
  16. 2024ColorPeel: Color Prompt Learning with Diffusion Models via Color and Shape Disentanglementcited by 28
  17. 2024Contrasting Deepfakes Diffusion via Contrastive Learning and Global-Local Similaritiescited by 27
  18. 2023Characteristic Guidance: Non-linear Correction for Diffusion Model at Large Guidance Scalecited by 26
  19. 2024LayerDiff: Exploring Text-guided Multi-layered Composable Image Synthesis via Layer-Collaborative Diffusion Modelcited by 26
  20. 2024MC2: Multi-concept Guidance for Customized Multi-concept Generationcited by 26
  21. 2024AID: Attention Interpolation of Text-to-Image Diffusioncited by 24
  22. 2023A Survey of Diffusion Based Image Generation Models: Issues and Their Solutionscited by 23
  23. 2024LiteVAE: Lightweight and Efficient Variational Autoencoders for Latent Diffusion Modelscited by 22
  24. 2024Self-Guidance: Boosting Flow and Diffusion Generation on Their Owncited by 22
  25. 2023Diffusion Model with Perceptual Losscited by 21
  26. 2023CLiC: Concept Learning in Contextcited by 20
  27. 2023Separate-and-Enhance: Compositional Finetuning for Text2Image Diffusion Modelscited by 19
  28. 2024Self-Guided Generation of Minority Samples Using Diffusion Modelscited by 19
  29. 2024MeshUp: Multi-Target Mesh Deformation via Blended Score Distillationcited by 17
  30. 2025Visual Generation Without Guidancecited by 17
  31. 2025Exploring Multimodal Diffusion Transformers for Enhanced Prompt-based Image Editingcited by 16
  32. 2023Style-A-Video: Agile Diffusion for Arbitrary Text-based Video Style Transfercited by 14
  33. 2024The Silent Assistant: NoiseQuery as Implicit Guidance for Goal-Driven Image Generationcited by 14
  34. 2025Token Perturbation Guidance for Diffusion Modelscited by 14
  35. 2026Diffusion Probe: Generated Image Result Prediction Using CNN Probescited by 13
  36. 2022DAG: Depth-Aware Guidance with Denoising Diffusion Probabilistic Modelscited by 12
  37. 2024PerLDiff: Controllable Street View Synthesis Using Perspective-Layout Diffusion Modelscited by 12
  38. 2025TCFG: Tangential Damping Classifier-free Guidancecited by 12

50 most cited of 151 All 151 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance
  2. 2023Elucidating The Design Space of Classifier-Guided Diffusion Generation
  3. 2024A Noise is Worth Diffusion Guidance
  4. 2023Adaptive Guidance: Training-free Acceleration of Conditional Diffusion Models
  5. 2022On Distillation of Guided Diffusion Models
  6. 2025HiGS: History-Guided Sampling for Plug-and-Play Enhancement of Diffusion Models
  7. 2023Bridging the Gap: Addressing Discrepancies in Diffusion Model Training for Classifier-Free Guidance
  8. 2022Self-Guided Diffusion Models