pith. sign in

arxiv: 2605.18736 · v1 · pith:KLWO234Znew · submitted 2026-05-18 · 💻 cs.CV

Spectral Progressive Diffusion for Efficient Image and Video Generation

Howard Xiao , Brian Chao , Lior Yariv , Gordon Wetzstein This is my paper

Pith reviewed 2026-05-20 11:03 UTC · model grok-4.3

classification 💻 cs.CV
keywords diffusion modelsefficient generationspectral methodsimage synthesisvideo generationprogressive resolutionfrequency domaindenoising trajectory
0
0 comments X p. Extension
pith:KLWO234Z Add to your LaTeX paper What is a Pith Number? 
\usepackage{pith}
\pithnumber{KLWO234Z}

Prints a linked pith:KLWO234Z badge after your title and writes the identifier into PDF metadata. Compiles on arXiv with no extra files. Learn more

The pith

Diffusion models for images and videos can run faster by progressively increasing resolution as denoising advances from low to high frequencies.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper shows that diffusion models already generate content autoregressively in frequency space, with low frequencies appearing first and high-frequency details later. This pattern makes full-resolution computation wasteful in the early, noise-heavy steps. The authors introduce a framework that starts generation at lower resolutions and expands them along the denoising path according to the model's own power spectrum. The method works on existing pretrained models without retraining and includes an optional fine-tuning step. If correct, it delivers measurable speed gains on current image and video generators while keeping output quality intact.

Core claim

We propose Spectral Progressive Diffusion, a general framework that progressively grows resolution along the denoising trajectory of pretrained diffusion models. We develop a spectral noise expansion mechanism and derive an optimal resolution schedule from the model's power spectrum. Our framework supports training-free acceleration and a novel fine-tuning recipe that further improves efficiency and quality. We demonstrate significant speedups on state-of-the-art pretrained image and video generation models while preserving visual quality.

What carries the argument

Spectral noise expansion mechanism paired with a resolution schedule derived from the model's power spectrum, which allows progressive growth of spatial resolution during the denoising trajectory.

If this is right

  • Pretrained diffusion models can generate images and videos with substantially lower compute cost without any retraining.
  • Quality metrics and human evaluations remain comparable to standard full-resolution runs because high-frequency detail is still synthesized at the appropriate later timesteps.
  • The same progressive schedule applies directly to both image and video generation models.
  • An optional fine-tuning stage can further reduce generation time or improve output fidelity beyond the training-free version.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach could be combined with existing sampling accelerators such as fewer steps or distillation to compound speed gains.
  • Similar progressive-resolution logic might transfer to other iterative generative models that operate across scales.
  • Lower per-sample compute could make high-resolution generation more practical on hardware with limited memory or power.

Load-bearing premise

High-resolution computation on noise-dominated frequencies is largely redundant and can be skipped without degrading the final visual quality.

What would settle it

A side-by-side comparison on the same pretrained model showing either no speedup, visible quality loss measured by standard metrics, or both when the progressive resolution schedule replaces full-resolution denoising from the start.

Figures

Figures reproduced from arXiv: 2605.18736 by Brian Chao, Gordon Wetzstein, Howard Xiao, Lior Yariv.

Figure 1
Figure 1. Figure 1: Spectral Progressive Diffusion. We progressively grow the resolution along the denoising trajectory using an optimal resolution schedule derived from the spectral power of pretrained models (left). At each scheduled transition, our spectral noise expansion mechanism (right) injects high￾frequency noise at the correct level while preserving the partially-denoised low-frequency content. denoised low-frequenc… view at source ↗
Figure 2
Figure 2. Figure 2: Diffusion process in the spectral domain. Latent power spectra in both image and video models decay rapidly with frequency (Fig. (a)), consistent with natural images. Diffusion exhibits a frequency-domain autoregressive structure (Fig. (b)) due to the aforementioned property: low frequencies emerge early in the denoising process, while high frequencies remain noise-dominated. 4 Spectral Progressive Diffusi… view at source ↗
Figure 3
Figure 3. Figure 3: Visual Generation Qualitative Comparisons. For the main comparison of latent-space image generation, our method outperforms the state-of-the-art spatial acceleration method RALU [32] in both visual fidelity and inference speed. Across all evaluated modalities (latent/pixel-space image generation and latent-space video generation), we achieve substantial acceleration over standard high-resolution baselines … view at source ↗
Figure 4
Figure 4. Figure 4: (a): Ablation Studies on δ, S and TΦ. We observe a clear tradeoff between image quality and efficiency when varying δ and S as shown in the top plot. Across transforms, DCT achieves similar quality as DWT and outperforms FFT as shown in the bottom plot. (b): Frequency-based Image Editing. Our method demonstrates superior prompt alignment and geometric consistency compared to standard SDEdit-style spatial-d… view at source ↗
Figure 5
Figure 5. Figure 5: Spectral noise passthrough experiment. At smaller δ ∈ [0.0001, 0.001], there is almost no observable difference compared to native full-resolution generation. As larger δ values cause high-frequency replacement to persist later in the denoising trajectory, we observe increasingly blurry and distorted results (i.e., ghosting artifacts and “CHOOLBUS” instead of “SCHOOLBUS”). 23 [PITH_FULL_IMAGE:figures/full… view at source ↗
Figure 6
Figure 6. Figure 6: Qualitative comparisons on latent-space image generation. We compare our method against default-step generation, reduced-step native-resolution generation on FLUX.1-dev [39], and RALU [32], a state-of-the-art acceleration baseline matched to similar speedups. Our method outperforms both baselines. FLUX.1-dev with reduced steps degrades image quality and exhibits over-saturation artifacts, while RALU introd… view at source ↗
Figure 7
Figure 7. Figure 7: Qualitative comparisons on latent-space image generation. We compare our method against default-step generation, reduced-step native-resolution generation on FLUX.1-dev [39], and RALU [32], a state-of-the-art acceleration baseline matched to similar speedups. Our method outperforms both baselines. FLUX.1-dev with reduced steps degrades image quality and exhibits over-saturation artifacts, while RALU introd… view at source ↗
Figure 8
Figure 8. Figure 8: Qualitative comparisons on latent-space image generation. We compare our method against default-step generation, reduced-step native-resolution generation on FLUX.1-dev [39], and RALU [32], a state-of-the-art acceleration baseline matched to similar speedups. Our method outperforms both baselines. FLUX.1-dev with reduced steps degrades image quality and exhibits over-saturation artifacts, while RALU introd… view at source ↗
Figure 9
Figure 9. Figure 9: Qualitative comparisons on latent-space image generation (fine-tuned). We compare our method against default-step generation, reduced-step native-resolution generation on Z-Image [5] matched to similar speedups. Our fine-tuned model (Ours∗ ) achieves even higher image quality compared to our training-free acceleration variant and outperforms the reduced-step baseline. 32 [PITH_FULL_IMAGE:figures/full_fig_… view at source ↗
Figure 10
Figure 10. Figure 10: Qualitative comparisons on latent-space image generation (fine-tuned). We compare our method against default-step generation, reduced-step native-resolution generation on Z-Image [5] matched to similar speedups. Our fine-tuned model (Ours∗ ) achieves even higher image quality compared to our training-free acceleration variant and outperforms the reduced-step baseline. 33 [PITH_FULL_IMAGE:figures/full_fig… view at source ↗
Figure 11
Figure 11. Figure 11: Qualitative comparisons on latent-space image generation (fine-tuned). We compare our method against default-step generation, reduced-step native-resolution generation on Z-Image [5] matched to similar speedups. Our fine-tuned model (Ours∗ ) achieves even higher image quality compared to our training-free acceleration variant and outperforms the reduced-step baseline. 34 [PITH_FULL_IMAGE:figures/full_fig… view at source ↗
Figure 12
Figure 12. Figure 12: Qualitative comparisons on pixel-space image generation. We compare our method against default-step generation and reduced-step native-resolution generation on PixelGen [55], matched to comparable speedups. An asterisk (Ours∗ ) marks the fine-tuned model. Our method achieves similar quality to full-resolution generation while attaining higher speedups. 35 [PITH_FULL_IMAGE:figures/full_fig_p035_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Qualitative comparisons on pixel-space image generation. We compare our method against default-step generation and reduced-step native-resolution generation on PixelGen [55], matched to comparable speedups. An asterisk (Ours∗ ) marks the fine-tuned model. Our method achieves similar quality to full-resolution generation while attaining higher speedups. 36 [PITH_FULL_IMAGE:figures/full_fig_p036_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Qualitative comparisons on pixel-space image generation. We compare our method against default-step generation and reduced-step native-resolution generation on PixelGen [55], matched to comparable speedups. An asterisk (Ours∗ ) marks the fine-tuned model. Our method achieves similar quality to full-resolution generation while attaining higher speedups. 37 [PITH_FULL_IMAGE:figures/full_fig_p037_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Qualitative ablation on TΦ. We see that FFT leads to overly smooth results while DCT and DWT attain similar image quality. 41 [PITH_FULL_IMAGE:figures/full_fig_p041_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Qualitative ablation on δ. We observe that increasing δ improves efficiency, but results in ghosting and halo artifacts near detailed edges. 42 [PITH_FULL_IMAGE:figures/full_fig_p042_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Qualitative ablation on S. We find that increasing S leads to marginal speedup improve￾ments and little image quality degradation. 43 [PITH_FULL_IMAGE:figures/full_fig_p043_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Texture editing results. Our frequency-based editing framework outperforms SDEdit, enabling high-fidelity texture transfer while preserving the geometric structure of the input image. 45 [PITH_FULL_IMAGE:figures/full_fig_p045_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Texture editing results. Our frequency-based editing framework outperforms SDEdit, enabling high-fidelity texture transfer while preserving the geometric structure of the input image. 46 [PITH_FULL_IMAGE:figures/full_fig_p046_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: Texture editing results. Our frequency-based editing framework outperforms SDEdit, enabling high-fidelity texture transfer while preserving the geometric structure of the input image. 47 [PITH_FULL_IMAGE:figures/full_fig_p047_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: Texture editing results. Our frequency-based editing framework outperforms SDEdit, enabling high-fidelity texture transfer while preserving the geometric structure of the input image. 48 [PITH_FULL_IMAGE:figures/full_fig_p048_21.png] view at source ↗
Figure 22
Figure 22. Figure 22: Artistic stylization results. Aside from texture editing, our frequency-based editing approach also supports artistic stylization given stylistic descriptions and a representative artist. 49 [PITH_FULL_IMAGE:figures/full_fig_p049_22.png] view at source ↗
Figure 23
Figure 23. Figure 23: Effect of TΦ on image editing. FFT–based editing leads to overly-smooth and hazy results; DCT- and DWT-based editing achieve similar editing quality. 50 [PITH_FULL_IMAGE:figures/full_fig_p050_23.png] view at source ↗
read the original abstract

Diffusion models have been shown to implicitly generate visual content autoregressively in the frequency domain, where low-frequency components are generated earlier in the denoising process while high-frequency details emerge only in later timesteps. This structure offers a natural opportunity for efficient generation, as high-resolution computation on noise-dominated frequencies is largely redundant. We propose Spectral Progressive Diffusion, a general framework that progressively grows resolution along the denoising trajectory of pretrained diffusion models. To this end, we develop a spectral noise expansion mechanism and derive an optimal resolution schedule from the model's power spectrum. Our framework supports training-free acceleration and a novel fine-tuning recipe that further improves efficiency and quality. We demonstrate significant speedups on state-of-the-art pretrained image and video generation models while preserving visual quality.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces Spectral Progressive Diffusion, a general framework for accelerating pretrained diffusion models in image and video generation. It progressively grows resolution along the denoising trajectory via a spectral noise expansion mechanism and derives an optimal resolution schedule from the model's power spectrum. The approach supports training-free acceleration as well as a fine-tuning recipe, with empirical claims of significant speedups on state-of-the-art models while preserving visual quality.

Significance. If the central efficiency claims hold under rigorous controls, the work could offer a practical, general-purpose acceleration technique for diffusion-based generation by exploiting the frequency-domain structure of the denoising process. The training-free component would be particularly valuable for immediate deployment on existing large models. The significance is reduced, however, by the need for stronger evidence that the schedule derivation is robust rather than circular or content-dependent.

major comments (2)
  1. [Methods (resolution schedule derivation)] The derivation of the resolution schedule from the model's power spectrum (detailed in the methods section on schedule construction) risks circularity: if the spectrum is averaged or fitted using statistics from the target model or dataset, the 'optimal' schedule becomes model-specific rather than an independent prediction. This directly affects the load-bearing claim of quality-preserving speedups. Please clarify the exact computation procedure, whether any fitting or selection occurs, and provide ablations showing performance on out-of-distribution content with atypical frequency distributions.
  2. [Experiments (qualitative and quantitative results)] The central efficiency claim relies on the premise that high-resolution computation on noise-dominated frequencies is redundant. However, the experiments appear to lack controls for content where high-frequency components carry semantic weight (e.g., fine textures or text). Without such targeted evaluations, the no-degradation guarantee remains unverified and undermines the abstract's assertion of preserved visual quality across SOTA models.
minor comments (2)
  1. [Section 3] Notation for the spectral noise expansion operator should be defined more explicitly with a clear equation reference to avoid ambiguity in the progressive growth description.
  2. [Figure 4] Figure captions for the resolution schedule plots should include the exact power spectrum computation details and any averaging window used.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed feedback on our manuscript. We address each major comment point by point below, providing clarifications on the methods and strengthening the experimental evidence where needed. Revisions have been made to incorporate additional details and results.

read point-by-point responses

  1. Referee: The derivation of the resolution schedule from the model's power spectrum (detailed in the methods section on schedule construction) risks circularity: if the spectrum is averaged or fitted using statistics from the target model or dataset, the 'optimal' schedule becomes model-specific rather than an independent prediction. This directly affects the load-bearing claim of quality-preserving speedups. Please clarify the exact computation procedure, whether any fitting or selection occurs, and provide ablations showing performance on out-of-distribution content with atypical frequency distributions.

    Authors: We thank the referee for raising this important issue of potential circularity. The power spectrum used for the schedule is computed by applying the pretrained diffusion model to a small fixed set of standard Gaussian noise inputs (independent of any target images or datasets) and averaging the frequency power magnitudes over timesteps to determine when high-frequency content emerges. No fitting, optimization, or content-based selection is performed; the resulting schedule reflects the model's inherent denoising behavior. We have expanded the methods section with this exact procedural description. Additionally, the revised manuscript includes new ablations on out-of-distribution content with atypical frequency distributions, such as text and synthetic high-frequency patterns, which confirm that the schedule generalizes while preserving quality. revision: yes

  2. Referee: The central efficiency claim relies on the premise that high-resolution computation on noise-dominated frequencies is redundant. However, the experiments appear to lack controls for content where high-frequency components carry semantic weight (e.g., fine textures or text). Without such targeted evaluations, the no-degradation guarantee remains unverified and undermines the abstract's assertion of preserved visual quality across SOTA models.

    Authors: We agree that targeted controls for semantically important high-frequency content are necessary to fully substantiate the no-degradation claim. Our original experiments used standard benchmarks containing varied textures and details, but we acknowledge the benefit of more focused evaluations. The revised paper now includes dedicated quantitative (FID, perceptual metrics) and qualitative results on images with fine textures, text, and complex patterns. These demonstrate that progressive resolution growth maintains semantic fidelity and visual quality equivalent to full-resolution baselines, as high-frequency details are introduced at appropriate later timesteps via spectral expansion. revision: yes

Circularity Check

0 steps flagged

Derivation of resolution schedule from power spectrum is independent and self-contained

full rationale

The paper states it develops a spectral noise expansion mechanism and derives an optimal resolution schedule from the model's power spectrum. No equations, sections, or self-citations are presented that reduce this derivation to a fitted parameter renamed as prediction, a self-definitional loop, or a load-bearing self-citation chain. The central efficiency claim rests on the empirical premise that high-resolution computation on noise-dominated frequencies is redundant, which is tested via demonstrations on pretrained models rather than forced by construction from the schedule itself. This is a standard first-principles analysis of frequency content in diffusion trajectories and does not exhibit the enumerated circularity patterns.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The approach rests on the observed frequency-generation ordering in diffusion models and on the ability to derive a resolution schedule from the power spectrum; no new physical entities are introduced.

free parameters (1)
  • resolution schedule thresholds
    Derived from the model's power spectrum; likely requires model-specific computation or selection to set growth points.
axioms (1)
  • domain assumption Diffusion models implicitly generate low-frequency components earlier than high-frequency details during denoising.
    Stated as the foundational observation enabling the progressive resolution approach.

pith-pipeline@v0.9.0 · 5651 in / 1188 out tokens · 37687 ms · 2026-05-20T11:03:22.366801+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

112 extracted references · 112 canonical work pages · 7 internal anchors

  1. [1]

    Ahmed, T

    N. Ahmed, T. Natarajan, and K. R. Rao. Discrete cosine transform.IEEE Transactions on Computers, C-23(1):90–93, 1974

    work page 1974

  2. [2]

    Longformer: The Long-Document Transformer

    Iz Beltagy, Matthew E Peters, and Arman Cohan. Longformer: The long-document transformer, 2020. URLhttps://arxiv.org/abs/2004.05150

    work page internal anchor Pith review Pith/arXiv arXiv 2020

  3. [3]

    Spectral analysis of diffusion models with application to schedule design

    Roi Benita, Miki Elad, and Joseph Keshet. Spectral analysis of diffusion models with application to schedule design. InAdv. Neural Inform. Process. Syst., volume 38, pages 2073–2127, 2026

    work page 2073

  4. [4]

    Token merging for fast stable diffusion

    Daniel Bolya and Judy Hoffman. Token merging for fast stable diffusion. InIEEE Conf. Comput. Vis. Pattern Recog., pages 4599–4603, 2023

    work page 2023

  5. [5]

    Z-Image: An efficient image generation foundation model with single-stream diffusion transformer, 2025

    Huanqia Cai, Sihan Cao, Ruoyi Du, Peng Gao, Steven Hoi, et al. Z-Image: An efficient image generation foundation model with single-stream diffusion transformer, 2025. URL https://arxiv.org/abs/2511. 22699

    work page 2025

  6. [6]

    Spectral regularization for diffusion models, 2026

    Satish Chandran, Nicolas Roque dos Santos, Yunshu Wu, Greg Ver Steeg, and Evangelos Papalexakis. Spectral regularization for diffusion models, 2026. URLhttps://arxiv.org/abs/2603.02447

    work page arXiv 2026

  7. [7]

    δ-DiT: A training-free acceleration method tailored for diffusion transformers, 2024

    Pengtao Chen, Mingzhu Shen, Peng Ye, Jianjian Cao, Chongjun Tu, Christos-Savvas Bouganis, Yiren Zhao, and Tao Chen. δ-DiT: A training-free acceleration method tailored for diffusion transformers, 2024. URLhttps://arxiv.org/abs/2406.01125

    work page arXiv 2024

  8. [8]

    Rethinking attention with performers

    Krzysztof Choromanski, Valerii Likhosherstov, David Dohan, Xingyou Song, Andreea Gane, Tamas Sarlos, Peter Hawkins, Jared Davis, Afroz Mohiuddin, Lukasz Kaiser, et al. Rethinking attention with performers. InInt. Conf. Learn. Represent., 2021

    work page 2021

  9. [9]

    Diffusion is spectral autoregression

    Sander Dieleman. Diffusion is spectral autoregression. Blog post, September 2024. URL https: //sander.ai/2024/09/02/spectral-autoregression.html

    work page 2024

  10. [10]

    DemoFusion: Democratising high-resolution image generation with no $$$

    Ruoyi Du, Dongliang Chang, et al. DemoFusion: Democratising high-resolution image generation with no $$$. InIEEE Conf. Comput. Vis. Pattern Recog., 2024

    work page 2024

  11. [11]

    Flow along the k-amplitude for generative modeling, 2025

    Weitao Du, Shuning Chang, Jiasheng Tang, Yu Rong, Fan Wang, and Shengchao Liu. Flow along the k-amplitude for generative modeling, 2025. URLhttps://arxiv.org/abs/2504.19353

    work page arXiv 2025

  12. [12]

    Scaling rectified flow transformers for high-resolution image synthesis

    Patrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari, Jonas Müller, Harry Saini, Yam Levi, Dominik Lorenz, Axel Sauer, Frederic Boesel, Dustin Podell, Tim Dockhorn, Zion English, and Robin Rombach. Scaling rectified flow transformers for high-resolution image synthesis. InInt. Conf. Mach. Learn., 2024

    work page 2024

  13. [13]

    Spectrally-guided diffusion noise schedules

    Carlos Esteves and Ameesh Makadia. Spectrally-guided diffusion noise schedules. InInt. Conf. Mach. Learn., 2026

    work page 2026

  14. [14]

    A fourier space perspective on diffusion models, 2025

    Fabian Falck, Teodora Pandeva, Kiarash Zahirnia, Rachel Lawrence, Richard Turner, Edward Meeds, Javier Zazo, and Sushrut Karmalkar. A fourier space perspective on diffusion models, 2025. URL https://arxiv.org/abs/2505.11278

    work page arXiv 2025

  15. [15]

    Vchitect-2.0: Parallel transformer for scaling up video diffusion models,

    Weichen Fan, Chenyang Si, Junhao Song, Zhenyu Yang, Yinan He, Long Zhuo, Ziqi Huang, Ziyue Dong, Jingwen He, Dongwei Pan, et al. Vchitect-2.0: Parallel transformer for scaling up video diffusion models,

    work page

  16. [16]

    URLhttps://arxiv.org/abs/2501.08453

    work page arXiv

  17. [17]

    Attend to not attended: Structure-then-detail token merging for post-training dit acceleration

    Haipeng Fang, Sheng Tang, Juan Cao, Enshuo Zhang, Fan Tang, and Tong-Yee Lee. Attend to not attended: Structure-then-detail token merging for post-training dit acceleration. InIEEE Conf. Comput. Vis. Pattern Recog., pages 18083–18092, 2025

    work page 2025

  18. [18]

    Firmin Didot, 1822

    Jean-Baptiste Joseph Fourier.Théorie Analytique de la Chaleur. Firmin Didot, 1822

    work page

  19. [19]

    GenEval: An object-focused framework for evaluating text-to-image alignment

    Dhruba Ghosh, Hannaneh Hajishirzi, and Ludwig Schmidt. GenEval: An object-focused framework for evaluating text-to-image alignment. InAdv. Neural Inform. Process. Syst., 2023

    work page 2023

  20. [20]

    Matryoshka diffusion models

    Jiatao Gu, Shuangfei Zhai, Yizhe Zhang, Joshua M Susskind, and Navdeep Jaitly. Matryoshka diffusion models. InInt. Conf. Learn. Represent., 2023

    work page 2023

  21. [21]

    Make a cheap scaling: A self-cascade diffusion model for higher-resolution adaptation

    Lanqing Guo, Yingqing He, Haoxin Chen, Menghan Xia, Xiaodong Cun, Yufei Wang, Siyu Huang, Yong Zhang, Xintao Wang, Qifeng Chen, Ying Shan, and Bihan Wen. Make a cheap scaling: A self-cascade diffusion model for higher-resolution adaptation. InEur. Conf. Comput. Vis., 2024. 11

    work page 2024

  22. [22]

    Wavelet score-based generative modeling

    Florentin Guth, Simon Coste, Valentin De Bortoli, and Stephane Mallat. Wavelet score-based generative modeling. InAdv. Neural Inform. Process. Syst., volume 35, pages 478–491, 2022

    work page 2022

  23. [23]

    Zur theorie der orthogonalen funktionensysteme.Mathematische Annalen, 69(3):331–371, 1910

    Alfréd Haar. Zur theorie der orthogonalen funktionensysteme.Mathematische Annalen, 69(3):331–371, 1910

    work page 1910

  24. [24]

    Infinity: Scaling bitwise AutoRegressive modeling for high-resolution image synthesis

    Jian Han, Jinlai Liu, Yi Jiang, Bin Yan, Yuqi Zhang, Zehuan Yuan, Bingyue Peng, and Xiaobing Liu. Infinity: Scaling bitwise AutoRegressive modeling for high-resolution image synthesis. InIEEE Conf. Comput. Vis. Pattern Recog., 2025

    work page 2025

  25. [25]

    Agglomerative token clustering

    Joakim Bruslund Haurum, Sergio Escalera, Graham W Taylor, and Thomas B Moeslund. Agglomerative token clustering. InEur. Conf. Comput. Vis., pages 200–218. Springer, 2024

    work page 2024

  26. [26]

    Denoising diffusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. InAdv. Neural Inform. Process. Syst., 2020

    work page 2020

  27. [27]

    Cascaded diffusion models for high fidelity image generation.J

    Jonathan Ho, Chitwan Saharia, William Chan, David J Fleet, Mohammad Norouzi, and Tim Salimans. Cascaded diffusion models for high fidelity image generation.J. Mach. Learn. Res., 23(47):1–33, 2022

    work page 2022

  28. [28]

    Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen

    Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. LoRA: Low-rank adaptation of large language models. InInt. Conf. Learn. Represent., 2022

    work page 2022

  29. [29]

    T2I-CompBench: A comprehensive benchmark for open-world compositional text-to-image generation

    Kaiyi Huang, Kaiyue Sun, Enze Xie, Zhenguo Li, and Xihui Liu. T2I-CompBench: A comprehensive benchmark for open-world compositional text-to-image generation. InAdv. Neural Inform. Process. Syst., 2023

    work page 2023

  30. [30]

    Wavedm: Wavelet-based diffusion models for image restoration.IEEE Trans

    Yi Huang, Jiancheng Huang, Jianzhuang Liu, Mingfu Yan, Yu Dong, Jiaxi Lv, Chaoqi Chen, and Shifeng Chen. Wavedm: Wavelet-based diffusion models for image restoration.IEEE Trans. Multimedia, 26: 7058–7073, 2024

    work page 2024

  31. [31]

    Spectralar: Spectral autoregressive visual generation

    Yuanhui Huang, Weiliang Chen, Wenzhao Zheng, Yueqi Duan, Jie Zhou, and Jiwen Lu. Spectralar: Spectral autoregressive visual generation. InInt. Conf. Comput. Vis., pages 15842–15852, 2025

    work page 2025

  32. [32]

    VBench: Comprehensive benchmark suite for video generative models

    Ziqi Huang, Yinan He, Jiashuo Yu, Fan Zhang, Chenyang Si, Yuming Jiang, Yuanhan Zhang, Tianxing Wu, Qingyang Jin, Nattapol Chanpaisit, Yaohui Wang, Xinyuan Chen, Limin Wang, Dahua Lin, Yu Qiao, and Ziwei Liu. VBench: Comprehensive benchmark suite for video generative models. InIEEE Conf. Comput. Vis. Pattern Recog., 2024

    work page 2024

  33. [33]

    Training-free mixed-resolution latent upsampling for spatially accelerated diffusion transformers, 2026

    Wongi Jeong, Kyungryeol Lee, Hoigi Seo, and Se Young Chun. Training-free mixed-resolution latent upsampling for spatially accelerated diffusion transformers, 2026. URLhttps://arxiv.org/abs/2507. 08422

    work page 2026

  34. [34]

    Pyramidal flow matching for efficient video generative modeling

    Yang Jin, Zhicheng Sun, Ningyuan Li, Kun Xu, Hao Jiang, Nan Zhuang, Quzhe Huang, Yang Song, Yadong Mu, and Zhouchen Lin. Pyramidal flow matching for efficient video generative modeling. InInt. Conf. Learn. Represent., 2025

    work page 2025

  35. [35]

    Kapfer, K

    C. Kapfer, K. Stine, B. Narasimhan, C. Mentzel, and E. Candès. Marlowe: Stanford’s GPU-based computational instrument, 2025. URLhttps://doi.org/10.5281/zenodo.14751899

    work page doi:10.5281/zenodo.14751899 2025

  36. [36]

    Transformers are rnns: Fast autoregressive transformers with linear attention

    Angelos Katharopoulos, Apoorv Vyas, Nikolaos Pappas, and François Fleuret. Transformers are rnns: Fast autoregressive transformers with linear attention. InInt. Conf. Mach. Learn., pages 5156–5165. PMLR, 2020

    work page 2020

  37. [37]

    Token fusion: Bridging the gap between token pruning and token merging

    Minchul Kim, Shangqian Gao, Yen-Chang Hsu, Yilin Shen, and Hongxia Jin. Token fusion: Bridging the gap between token pruning and token merging. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 1383–1392, 2024

    work page 2024

  38. [38]

    DiffuseHigh: Training-free progressive high-resolution image synthesis through structure guidance

    Younghyun Kim, Geunmin Hwang, Junyu Zhang, and Eunbyung Park. DiffuseHigh: Training-free progressive high-resolution image synthesis through structure guidance. InAAAI, 2025

    work page 2025

  39. [39]

    Kingma, Tim Salimans, Ben Poole, and Jonathan Ho

    Diederik P. Kingma, Tim Salimans, Ben Poole, and Jonathan Ho. Variational diffusion models. InAdv. Neural Inform. Process. Syst., 2021

    work page 2021

  40. [40]

    Black Forest Labs. FLUX. Software repository, 2024. URL https://github.com/ black-forest-labs/flux

    work page 2024

  41. [41]

    FLUX.2: Frontier Visual Intelligence

    Black Forest Labs. FLUX.2: Frontier Visual Intelligence. Blog post, 2025. URL https://bfl.ai/ blog/flux-2. 12

    work page 2025

  42. [42]

    Beta Sampling is All You Need: Efficient Image Generation Strategy for Diffusion Models using Stepwise Spectral Analysis

    Haeil Lee, Hansang Lee, Seoyeon Gye, and Junmo Kim. Beta sampling is all you need: Efficient image generation strategy for diffusion models using stepwise spectral analysis, 2024. URL https: //arxiv.org/abs/2407.12173

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  43. [43]

    Local representative token guided merging for text-to-image generation, 2025

    Min-Jeong Lee, Hee-Dong Kim, and Seong-Whan Lee. Local representative token guided merging for text-to-image generation, 2025. URLhttps://arxiv.org/abs/2507.12771

    work page arXiv 2025

  44. [44]

    Frecas: Efficient higher-resolution image generation via frequency-aware cascaded sampling

    Ruihuang Li, Lei Zhang, et al. Frecas: Efficient higher-resolution image generation via frequency-aware cascaded sampling. InInt. Conf. Learn. Represent., volume 2025, pages 6400–6412, 2025

    work page 2025

  45. [45]

    Radial attention: O(nlogn) sparse attention with energy decay for long video generation

    Xingyang Li, Muyang Li, Tianle Cai, Haocheng Xi, Shuo Yang, Yujun Lin, Lvmin Zhang, Songlin Yang, Jinbo Hu, Kelly Peng, et al. Radial attention: O(nlogn) sparse attention with energy decay for long video generation. InAdv. Neural Inform. Process. Syst., 2025

    work page 2025

  46. [46]

    Wf-vae: Enhancing video vae by wavelet-driven energy flow for latent video diffusion model

    Zongjian Li, Bin Lin, Yang Ye, Liuhan Chen, Xinhua Cheng, Shenghai Yuan, and Li Yuan. Wf-vae: Enhancing video vae by wavelet-driven energy flow for latent video diffusion model. InIEEE Conf. Comput. Vis. Pattern Recog., pages 17778–17788, 2025

    work page 2025

  47. [47]

    Lawrence Zitnick

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C. Lawrence Zitnick. Microsoft COCO: Common objects in context. InEur. Conf. Comput. Vis., 2014

    work page 2014

  48. [48]

    Yaron Lipman, Ricky T. Q. Chen, Heli Ben-Hamu, Maximilian Nickel, and Matt Le. Flow matching for generative modeling. InInt. Conf. Learn. Represent., 2023

    work page 2023

  49. [49]

    Timestep embedding tells: It’s time to cache for video diffusion model.arXiv preprint arXiv:2411.19108, 2024

    Feng Liu, Shiwei Zhang, Xiaofeng Wang, Yujie Wei, Haonan Qiu, Yuzhong Zhao, Yingya Zhang, Qixiang Ye, and Fang Wan. Timestep embedding tells: It’s time to cache for video diffusion model, 2024. URL https://arxiv.org/abs/2411.19108

    work page arXiv 2024

  50. [50]

    FreqCa: Accelerating diffusion models via frequency-aware caching, 2025

    Jiacheng Liu, Peiliang Cai, Qinming Zhou, Yuqi Lin, Deyang Kong, Benhao Huang, Yupei Pan, Haowen Xu, Chang Zou, Junshu Tang, Shikang Zheng, and Linfeng Zhang. FreqCa: Accelerating diffusion models via frequency-aware caching, 2025. URLhttps://arxiv.org/abs/2510.08669

    work page arXiv 2025

  51. [51]

    From reusing to forecasting: Accelerating diffusion models with TaylorSeers

    Jiacheng Liu, Chang Zou, Yuanhuiyi Lyu, Junjie Chen, and Linfeng Zhang. From reusing to forecasting: Accelerating diffusion models with TaylorSeers. InInt. Conf. Comput. Vis., 2025

    work page 2025

  52. [52]

    Flow straight and fast: Learning to generate and transfer data with rectified flow

    Xingchao Liu, Chengyue Gong, and Qiang Liu. Flow straight and fast: Learning to generate and transfer data with rectified flow. InInt. Conf. Learn. Represent., 2023

    work page 2023

  53. [53]

    ToMA: Token merge with attention for diffusion models

    Wenbo Lu, Shaoyi Zheng, Yuxuan Xia, and Shengjie Wang. ToMA: Token merge with attention for diffusion models. InInt. Conf. Mach. Learn., 2025

    work page 2025

  54. [54]

    Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference

    Simian Luo, Yiqin Tan, Longbo Huang, Jian Li, and Hang Zhao. Latent consistency models: Synthesizing high-resolution images with few-step inference, 2023. URLhttps://arxiv.org/abs/2310.04378

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  55. [55]

    Deepcache: Accelerating diffusion models for free

    Xinyin Ma, Gongfan Fang, and Xinchao Wang. Deepcache: Accelerating diffusion models for free. In IEEE Conf. Comput. Vis. Pattern Recog., 2024

    work page 2024

  56. [56]

    PixelGen: Improving Pixel Diffusion with Perceptual Supervision

    Zehong Ma, Ruihan Xu, and Shiliang Zhang. PixelGen: Pixel diffusion beats latent diffusion with perceptual loss, 2026. URLhttps://arxiv.org/abs/2602.02493

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  57. [57]

    SDEdit: Guided image synthesis and editing with stochastic differential equations

    Chenlin Meng, Yutong He, Yang Song, Jiaming Song, Jiajun Wu, Jun-Yan Zhu, and Stefano Ermon. SDEdit: Guided image synthesis and editing with stochastic differential equations. InInt. Conf. Learn. Represent., 2022

    work page 2022

  58. [58]

    completely blind

    Anish Mittal, Rajiv Soundararajan, and Alan C. Bovik. Making a “completely blind” image quality analyzer.IEEE Sign. Process. Letters, 20(3):209–212, 2013. doi: 10.1109/LSP.2012.2227726

    work page doi:10.1109/lsp.2012.2227726 2013

  59. [59]

    Scale space diffusion, 2026

    Soumik Mukhopadhyay, Prateksha Udhayanan, and Abhinav Shrivastava. Scale space diffusion, 2026. URLhttps://arxiv.org/abs/2603.08709

    work page arXiv 2026

  60. [60]

    DCTdiff: Intriguing properties of image generative modeling in the DCT space

    Mang Ning, Mingxiao Li, Jianlin Su, Jia Haozhe, Lanmiao Liu, Martin Benes, Wenshuo Chen, Albert Ali Salah, and Itir Onal Ertugrul. DCTdiff: Intriguing properties of image generative modeling in the DCT space. InInt. Conf. Mach. Learn., volume 267, pages 46498–46524. PMLR, 2025

    work page 2025

  61. [61]

    NVIDIA, Yuval Atzmon, Maciej Bala, Yogesh Balaji, Tiffany Cai, Yin Cui, Jiaojiao Fan, Yunhao Ge, Siddharth Gururani, Jacob Huffman, Ronald Isaac, Pooya Jannaty, Tero Karras, Grace Lam, J. P. Lewis, Aaron Licata, Yen-Chen Lin, Ming-Yu Liu, Qianli Ma, Arun Mallya, Ashlee Martino-Tarr, Doug Mendez, Seungjun Nah, Chris Pruett, Fitsum Reda, Jiaming Song, Ting-...

    work page arXiv 2024

  62. [62]

    Certain topics in telegraph transmission theory.Transactions of the American Institute of Electrical Engineers, 47(2):617–644, 1928

    Harry Nyquist. Certain topics in telegraph transmission theory.Transactions of the American Institute of Electrical Engineers, 47(2):617–644, 1928

    work page 1928

  63. [63]

    Scalable diffusion models with transformers

    William Peebles and Saining Xie. Scalable diffusion models with transformers. InInt. Conf. Comput. Vis., 2023

    work page 2023

  64. [64]

    Wavelet diffusion models are fast and scalable image generators

    Hao Phung, Quan Dao, and Anh Tran. Wavelet diffusion models are fast and scalable image generators. In IEEE Conf. Comput. Vis. Pattern Recog., 2023

    work page 2023

  65. [65]

    DiMSUM: Diffusion Mamba – a scalable and unified spatial-frequency method for image generation

    Hao Phung, Quan Dao, Trung Dao, Hoang Phan, Dimitris Metaxas, and Anh Tran. DiMSUM: Diffusion Mamba – a scalable and unified spatial-frequency method for image generation. InAdv. Neural Inform. Process. Syst., 2024

    work page 2024

  66. [66]

    FreeScale: Unleashing the resolution of diffusion models via tuning-free scale fusion

    Haonan Qiu et al. FreeScale: Unleashing the resolution of diffusion models via tuning-free scale fusion. In Int. Conf. Comput. Vis., 2025

    work page 2025

  67. [67]

    FlowAR: Scale-wise autoregressive image generation meets flow matching

    Sucheng Ren, Qihang Yu, Ju He, Xiaohui Shen, Alan Yuille, and Liang-Chieh Chen. FlowAR: Scale-wise autoregressive image generation meets flow matching. InInt. Conf. Mach. Learn., 2025

    work page 2025

  68. [68]

    Origins of scaling in natural images.Vis

    Daniel L Ruderman. Origins of scaling in natural images.Vis. Res., 37(23):3385–3398, 1997

    work page 1997

  69. [69]

    Pyramidal denoising diffusion probabilistic models, 2022

    Dohoon Ryu and Jong Chul Ye. Pyramidal denoising diffusion probabilistic models, 2022. URL https: //arxiv.org/abs/2208.01864

    work page arXiv 2022

  70. [70]

    Progressive distillation for fast sampling of diffusion models

    Tim Salimans and Jonathan Ho. Progressive distillation for fast sampling of diffusion models. InInt. Conf. Learn. Represent., 2022

    work page 2022

  71. [71]

    Claude E. Shannon. Communication in the presence of noise.Proceedings of the IRE, 37(1):10–21, 1949

    work page 1949

  72. [72]

    Latent Wavelet Diffusion For Ultra-High-Resolution Image Synthesis

    Luigi Sigillo, Shengfeng He, and Danilo Comminiello. Latent wavelet diffusion for ultra-high-resolution image synthesis, 2025. URLhttps://arxiv.org/abs/2506.00433

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  73. [73]

    Improving the diffusability of autoencoders

    Ivan Skorokhodov, Sharath Girish, Benran Hu, Willi Menapace, Yanyu Li, Rameen Abdal, Sergey Tulyakov, and Aliaksandr Siarohin. Improving the diffusability of autoencoders. InInt. Conf. Mach. Learn., 2025

    work page 2025

  74. [74]

    Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole

    Yang Song, Jascha Sohl-Dickstein, Diederik P. Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. Score-based generative modeling through stochastic differential equations. InInt. Conf. Learn. Represent., 2021

    work page 2021

  75. [75]

    Lssgen: Leveraging latent space scaling in flow and diffusion for efficient text to image generation

    Jyun-Ze Tang, Chih-Fan Hsu, Jeng-Lin Li, Ming-Ching Chang, and Wei-Chao Chen. Lssgen: Leveraging latent space scaling in flow and diffusion for efficient text to image generation. InInt. Conf. Comput. Vis., pages 5048–5057, 2025

    work page 2025

  76. [77]

    Relay diffusion: Unifying diffusion process across resolutions for image synthesis

    Jiayan Teng, Wendi Zheng, Ming Ding, Wenyi Hong, Jianqiao Wangni, Zhuoyi Yang, and Jie Tang. Relay diffusion: Unifying diffusion process across resolutions for image synthesis. InInt. Conf. Learn. Represent.,

    work page

  77. [78]

    Visual autoregressive modeling: Scalable image generation via next-scale prediction

    Keyu Tian, Yi Jiang, Zehuan Yuan, Bingyue Peng, and Liwei Wang. Visual autoregressive modeling: Scalable image generation via next-scale prediction. InAdv. Neural Inform. Process. Syst., 2024

    work page 2024

  78. [79]

    Training-free diffusion acceleration with bottleneck sampling, 2025

    Ye Tian, Xin Xia, Yuxi Ren, Shanchuan Lin, Xing Wang, Xuefeng Xiao, Yunhai Tong, Ling Yang, and Bin Cui. Training-free diffusion acceleration with bottleneck sampling, 2025. URL https://arxiv.org/ abs/2503.18940

    work page arXiv 2025

  79. [80]

    Hiwave: Training-free high- resolution image generation via wavelet-based diffusion sampling

    Tobias V ontobel, Seyedmorteza Sadat, Farnood Salehi, and Romann Weber. Hiwave: Training-free high- resolution image generation via wavelet-based diffusion sampling. InProceedings of the SIGGRAPH Asia 2025 Conference Papers, pages 1–11, 2025

    work page 2025

  80. [81]

    Wan: Open and Advanced Large-Scale Video Generative Models

    Team Wan, Ang Wang, Baole Ai, Bin Wen, Chaojie Mao, Chen-Wei Xie, Di Chen, Feiwu Yu, Haiming Zhao, Jianxiao Yang, et al. Wan: Open and advanced large-scale video generative models, 2025. URL https://arxiv.org/abs/2503.20314

    work page internal anchor Pith review Pith/arXiv arXiv 2025

Showing first 80 references.