提出热消散流匹配，用模糊过程提升图像生成多尺度细节。
Multi-Scale Generative Modeling with Heat Dissipation Flow Matching
- 将热消散过程引入流匹配框架，注入多尺度先验。
- 在多个数据集上优于主流基线方法，尤其在保持颜色和细节上表现更佳。
- 适合关注图像生成质量与多尺度建模的研究者。
扩散模型广泛用于图像生成，多数依赖噪声污染与去噪；另一分支则以模糊作为主要污染方式，通过提供多尺度先验，更好地保留颜色预算和多尺度细节。然而，基于模糊的模型仍局限于SDE框架，未融入如流匹配（FM）的ODE框架。此外，在数据流形假设下，从高维噪声（或速度）空间回归模糊图像面临困难。本文提出热消散流匹配（HDFM），将连续模糊（热消散）过程引入流匹配，注入多尺度先验。HDFM通过对齐插值热消散路径解决病态问题，并采用x-预测缓解高维回归难题。玩具实验与消融研究显示，HDFM在模糊和x-预测上均持续获益。在所有数据集上，HDFM性能超越多数基线方法。
原文摘要 · Abstract (English)
Diffusion models are widely used in image generation, with most relying on noise-based corruption and denoising. A distinct branch instead uses blur as the main corruption, preserving better color budgets and multi-scale detail by providing multi-scale priors. However, blur-based models remain in SDE-based frameworks and are not integrated into ODE-based frameworks, such as Flow Matching (FM). Meanwhile, in the blur-based formulation, the classical inverse heat-dissipation (IHD) process faces an ill-posed challenge. Moreover, under the data-manifold assumption, regressing blurred images from high-dimensional noise (or velocity) space is also difficult. We propose Heat Dissipation Flow Matching (HDFM), which introduces a continuous blurred (heat-dissipation) process into FM to inject multi-scale priors. HDFM aligns an interpolated heat-dissipation path to address ill-posedness and adopts $x$-prediction to mitigate high-dimensional regression difficulty. Toy experiments and ablation studies show that HDFM consistently benefits from both blur and $x$-prediction. The performance of HDFM outperforms most baseline methods on all datasets.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。