mlsanityData to 7 Oct 2026

I-Max: Maximize the Resolution Potential of Pre-trained Rectified Flow Transformers with Projected Flow

arXiv 2410.07536 ↗First posted 10 Oct 2024cs.CVcited by 21

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Rectified Flow Transformers (RFTs) offer superior training and inference efficiency, making them likely the most viable direction for scaling up diffusion models. However, progress in generation resolution has been relatively slow due to data quality and training costs. Tuning-free resolution extrapolation presents an alternative, but current methods often reduce generative stability, limiting practical application. In this paper, we review existing resolution extrapolation methods and introduce the I-Max framework to maximize the resolution potential of Text-to-Image RFTs. I-Max features: (i) a novel Projected Flow strategy for stable extrapolation and (ii) an advanced inference toolkit for generalizing model knowledge to higher resolutions. Experiments with Lumina-Next-2K and Flux.1-dev demonstrate I-Max's ability to enhance stability in resolution extrapolation and show that it can bring image detail emergence and artifact correction, confirming the practical value of tuning-free resolution extrapolation.

Key facts. Cited by 21 papers in the map (data through 7 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 36 of its 41 references that are in the map. First posted on arXiv 10 Oct 2024.

Builds on

36 of 41 references.

Closest ancestors · 22

Same direction, newest first

  1. 2024Lumina-Next: Making Lumina-T2X Stronger and Faster with Next-DiT5 Jun
  2. 2024Lumina-T2X: Transforming Text into Any Modality, Resolution, and Duration via Flow-based Large Diffusion Transformers9 May
  3. 2024CutDiffusion: A Simple, Fast, Cheap, and Strong Diffusion Extrapolation Method23 Apr
  4. 2024Upsample Guidance: Scale Up Diffusion Models without Training2 Apr
  5. 2024FouriScale: A Frequency Perspective on Training-Free High-Resolution Image Synthesis19 Mar
  6. 2024PixArt-Sigma: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generation7 Mar
  7. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesis5 Mar
  8. 2024ResAdapter: Domain Consistent Resolution Adapter for Diffusion Models4 Mar
  9. 2024FiT: Flexible Vision Transformer for Diffusion Model19 Feb
  10. 2024SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers16 Jan
  11. 2023DeepCache: Accelerating Diffusion Models for Free1 Dec
  12. 2023ScaleCrafter: Tuning-free Higher-Resolution Visual Generation with Diffusion Models11 Oct
10 more
  1. 2023Relay Diffusion: Unifying diffusion process across resolutions for image synthesis4 Sep
  2. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis4 Jul
  3. 2022Scalable Diffusion Models with Transformers19 Dec
  4. 2022Classifier-Free Diffusion Guidance26 Jul
  5. 2022Perception Prioritized Training of Diffusion Models1 Apr
  6. 2021High-Resolution Image Synthesis with Latent Diffusion Models20 Dec
  7. 2021Cascaded Diffusion Models for High Fidelity Image Generation30 May
  8. 2020Score-Based Generative Modeling through Stochastic Differential Equations26 Nov
  9. 2020Denoising Diffusion Implicit Models6 Oct
  10. 2020Denoising Diffusion Probabilistic Models19 Jun

Other references · 14

  1. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,005
  2. 2022Flow Matching for Generative Modelingcited by 5,426
  3. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,581
  4. 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,550
  5. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,928
  6. 2021SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equationscited by 1,903
  7. 2018ESRGAN: Enhanced Super-Resolution Generative Adversarial Networkscited by 1,250
  8. 2022Token Merging: Your ViT But Fastercited by 995
  9. 2023MultiDiffusion: Fusing Diffusion Paths for Controlled Image Generationcited by 529
  10. 2023DemoFusion: Democratising High-Resolution Image Generation With No cited by 105
  11. 2024UltraPixel: Advancing Ultra-High-Resolution Image Synthesis to New Peakscited by 64
  12. 2023Training-free Diffusion Model Adaptation for Variable-Sized Text-to-Image Synthesiscited by 55
2 more
  1. 2023ElasticDiffusion: Training-free Arbitrary Size Image Generation through Global-Local Content Separationcited by 37
  2. 2024AccDiffusion: An Accurate Method for Higher-Resolution Image Generationcited by 32

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 9; 2026: 10.

21 papers.

Sort:

By area:

  1. 2024FreeScale: Unleashing the Resolution of Diffusion Models via Tuning-Free Scale Fusioncited by 29
  2. 2024CLEAR: Conv-Like Linearization Revs Pre-Trained Diffusion Transformers Upcited by 26
  3. 2025HiFlow: Training-free High-Resolution Image Generation with Flow-Aligned Guidancecited by 21
  4. 2025UltraHR-100K: Enhancing UHR Image Synthesis with A Large-Scale High-Quality Datasetcited by 16
  5. 2025DyPE: Dynamic Position Extrapolation for Ultra High Resolution Diffusioncited by 13
  6. 2025UltraImage: Rethinking Resolution Extrapolation in Image Diffusion Transformerscited by 10
  7. 2025HiStream: Efficient High-Resolution Video Generation via Redundancy-Eliminated Streamingcited by 8
  8. 2026L2P: Unlocking Latent Potential for Pixel Generationcited by 7
  9. 2025Latent Wavelet Diffusion For Ultra-High-Resolution Image Synthesiscited by 6
  10. 2025FreeSwim: Revisiting Sliding-Window Attention Mechanisms for Training-Free Ultra-High-Resolution Video Generationcited by 6
  11. 2026ViBe: Ultra-High-Resolution Video Synthesis Born from Pure Imagescited by 2
  12. 2026SEGA: Spectral-Energy Guided Attention for Resolution Extrapolation in Diffusion Transformerscited by 2
9 more
  1. 2025UltraImageGen: Efficient Ultra-High-Resolution Image Generation with Hierarchical Local Attentioncited by 1
  2. 2025ResDiT: Evoking the Intrinsic Resolution Scalability in Diffusion Transformerscited by 1
  3. 2026PixVerve: Advancing Native UHR Image Generation to 100MP with a Large-Scale High-Quality Datasetcited by 1
  4. 2026ExtraVAR: Stage-Aware RoPE Remapping for Resolution Extrapolation in Visual Autoregressive Modelscited by 0
  5. 2026HierEdit: Region-Aware Hierarchical Diffusion for Efficient High-Resolution Editingcited by 0
  6. 2026VINS-120K: Ultra High-Resolution Image Editing with A Large-Scale Datasetcited by 0
  7. 2026Tuning-Free Latent Diffusion Models for Ultrahigh-Resolution Image Editingcited by 0
  8. 2026HRDiT: Training-Free High-Resolution Image Generation with Off-the-Shelf Diffusion Transformer Modelscited by 0
  9. 2026EDITBRIDGE: Towards Faithful and Efficient Ultra-High-Resolution Image Editingcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Multi-Resolution Flow Matching: Training-Free Diffusion Acceleration via Staged Sampling
  2. 2026Reflective Flow Sampling Enhancement
  3. 2026TIDE: Text-Informed Dynamic Extrapolation with Step-Aware Temperature Control for Diffusion Transformers
  4. 2024MegaFusion: Extend Diffusion Models towards Higher-resolution Image Generation without Further Tuning
  5. 2023Boosting Latent Diffusion with Flow Matching
  6. 2024Text-to-Image Rectified Flow as Plug-and-Play Priors
  7. 2026Fast, faithful and photorealistic diffusion-based image super-resolution with enhanced Flow Map models
  8. 2026RFMSR: Residual Flow Matching for Image Super-Resolution