mlsanityData to 7 Oct 2026

SenseFlow: Scaling Distribution Matching for Flow-based Text-to-Image Distillation

arXiv 2506.00523 ↗First posted 31 May 2025cs.CVcited by 29

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

The Distribution Matching Distillation (DMD) has been successfully applied to text-to-image diffusion models such as Stable Diffusion (SD) 1.5. However, vanilla DMD suffers from convergence difficulties on large-scale flow-based text-to-image models, such as SD 3.5 and FLUX. In this paper, we first analyze the issues when applying vanilla DMD on large-scale models. Then, to overcome the scalability challenge, we propose implicit distribution alignment (IDA) to constrain the divergence between the generator and the fake distribution. Furthermore, we propose intra-segment guidance (ISG) to relocate the timestep denoising importance from the teacher model. With IDA alone, DMD converges for SD 3.5; employing both IDA and ISG, DMD converges for SD 3.5 and FLUX.1 dev. Together with a scaled VFM-based discriminator, our final model, dubbed SenseFlow, achieves superior performance in distillation for both diffusion based text-to-image models such as SDXL, and flow-matching models such as SD 3.5 Large and FLUX.1 dev. The source code is available at hrefhttps://github.com/XingtongGe/SenseFlowhttps://github.com/XingtongGe/SenseFlow

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 39 of its 48 references that are in the map. First posted on arXiv 31 May 2025.

Builds on

39 of 48 references.

Closest ancestors · 20

Same direction, newest first

  1. 2025RayFlow: Instance-Aware Diffusion Acceleration via Adaptive Flow Trajectories10 Mar
  2. 2024Simplifying, Stabilizing and Scaling Continuous-Time Consistency Models14 Oct
  3. 2024Flash Diffusion: Accelerating Any Conditional Diffusion Model for Few Steps Image Generation4 Jun
  4. 2024Phased Consistency Models28 May
  5. 2024Improved Distribution Matching Distillation for Fast Image Synthesis23 May
  6. 2024Hyper-SD: Trajectory Segmented Consistency Model for Efficient Image Synthesis21 Apr
  7. 2024You Only Sample Once: Taming One-Step Text-to-Image Synthesis by Self-Cooperative Diffusion GANs19 Mar
  8. 2024Fast High-Resolution Image Synthesis with Latent Adversarial Diffusion Distillation18 Mar
  9. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesis5 Mar
  10. 2024SDXL-Lightning: Progressive Adversarial Diffusion Distillation21 Feb
  11. 2023One-step Diffusion with Distribution Matching Distillation30 Nov
  12. 2023Adversarial Diffusion Distillation28 Nov
8 more
  1. 2023Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference6 Oct
  2. 2023Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion1 Oct
  3. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis4 Jul
  4. 2023Diff-Instruct: A Universal Approach for Transferring Knowledge From Pre-trained Diffusion Models29 May
  5. 2023Consistency Models2 Mar
  6. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  7. 2021Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed7 Jan
  8. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 19

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  3. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  4. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  5. 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,131
  6. 2022DreamFusion: Text-to-3D using 2D Diffusioncited by 3,006
  7. 2024SAM 2: Segment Anything in Images and Videoscited by 2,784
  8. 2023ProlificDreamer: High-Fidelity and Diverse Text-to-3D Generation with Variational Score Distillationcited by 1,207
  9. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,010
  10. 2023Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generationcited by 950
  11. 2023GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignmentcited by 946
  12. 2023Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesiscited by 847
7 more
  1. 2023Diffusion Model Alignment Using Direct Preference Optimizationcited by 767
  2. 2022Diffusion-GAN: Training GANs with Diffusioncited by 203
  3. 2023AM-RADIO: Agglomerative Vision Foundation Model -- Reduce All Domains Into Onecited by 165
  4. 2024Playground v3: Improving Text-to-Image Alignment with Deep-Fusion Large Language Modelscited by 113
  5. 2025SANA-Sprint: One-Step Diffusion with Continuous-Time Consistency Distillationcited by 85
  6. 2025Seedream 2.0: A Native Chinese-English Bilingual Image Generation Foundation Modelcited by 59
  7. 2024Efficient Video Face Enhancement with Enhanced Spatial-Temporal Consistencycited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 7; 2026: 22.

29 papers.

Sort:

By area:

  1. 2025Distribution Matching Distillation Meets Reinforcement Learningcited by 43
  2. 2025pi-Flow: Policy-Based Few-Step Generation via Imitation Distillationcited by 25
  3. 2025Flash-DMD: Towards High-Fidelity Few-Step Image Generation with Efficient Distillation and Joint Reinforcement Learningcited by 17
  4. 2026Salt: Self-Consistent Distribution Matching with Cache-Aware Training for Fast Video Generationcited by 11
  5. 2025LinVideo: A Post-Training Framework towards O(n) Attention in Efficient Video Generationcited by 9
  6. 2025LiveTalk: Real-Time Multimodal Interactive Video Diffusion via Improved On-Policy Distillationcited by 8
  7. 2026Reinforcing Few-step Generators via Reward-Tilted Distribution Matchingcited by 6
  8. 2026Flash-VAED: Plug-and-Play VAE Decoders for Efficient Video Generationcited by 5
  9. 2025VDOT: Efficient Unified Video Creation via Optimal Transport Distillationcited by 4
  10. 2026Continuous-Time Distribution Matching for Few-Step Diffusion Distillationcited by 4
  11. 2026Distribution Matching Distillation without Fake Score Networkcited by 4
  12. 2026Lens: Rethinking Training Efficiency for Foundational Text-to-Image Modelscited by 4
17 more
  1. 2025Sortblock: Similarity-Aware Feature Reuse for Diffusion Modelcited by 3
  2. 2026Mean Flow Distillation: Robust and Stable Distillation for Flow Matching Modelscited by 3
  3. 2026Budget-Constrained Step-Level Diffusion Cachingcited by 2
  4. 2026Multi-Resolution Flow Matching: Training-Free Diffusion Acceleration via Staged Samplingcited by 2
  5. 2026DefaultShift: Auditing Semantic Default Shift in Accelerated Text-to-Image Modelscited by 2
  6. 2026Reward-Aware Trajectory Shaping for Few-step Visual Generationcited by 1
  7. 2026Omni-LiveAvatar: Minute-Level Real-Time Streaming Joint Audio-Video Avatar Generationcited by 1
  8. 2026CrossDistill: Balancing Quality and Diversity via Trajectory-Level Hybrid Few-Step Distillationcited by 1
  9. 2026Salt++: Context-Aligned Post-Training for Few-Step Streaming Multimodal Generationcited by 1
  10. 2026IB-Flow: Information Bottleneck-Guided CFG Distillation for Few-Step Text-to-Image Generationcited by 0
  11. 2026Spectral Amplitude Purification in Distribution Matching for Diffusion Distillationcited by 0
  12. 2026PDMD: Projected Distribution Matching Distillation for Video Diffusion Modelscited by 0
  13. 2026Waypoint-1.5: A Real-Time Video World Model for Consumer Hardwarecited by 0
  14. 2026SoL-Refiner: Speed-of-Light One-Step Refinement for High-Resolution Videocited by 0
  15. 2026Enhancing Autoregressive Video Generation via Representation Adversarial Distillationcited by 0
  16. 2026DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generationcited by 0
  17. 2026Multi-Bandwidth Distribution Matching Distillation: On the Equivalence of Distribution Matching Distillation and Drifting Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 20261.x-Distill: Breaking the Diversity, Quality, and Efficiency Barrier in Distribution Matching Distillation
  2. 2025Score Distillation of Flow Matching Models
  3. 2025Few-Step Diffusion via Score identity Distillation
  4. 2026DMGD: Train-Free Dataset Distillation with Semantic-Distribution Matching in Diffusion Models
  5. 2024Regularized Distribution Matching Distillation for One-step Unpaired Image-to-Image Translation
  6. 2025Scale-wise Distillation of Diffusion Models
  7. 2026Better Source, Better Flow: Learning Condition-Dependent Source Distribution for Flow Matching
  8. 2025CoDA: From Text-to-Image Diffusion Models to Training-Free Dataset Distillation