arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2409.03755cs.CV
收藏

DC-Solver:通过动态补偿改进预测-校正扩散采样器

DC-Solver: Improving Predictor-Corrector Diffusion Sampler via Dynamic Compensation

Tsinghua University(清华大学);Beijing National Research Center for Information Science and Technology(北京信息科学与技术国家研究中心)

Wenliang Zhao, Haolin Wang, Jie Zhou, Jiwen Lu

AI总结:

本文提出DC-Solver,通过动态补偿和级联多项式回归缓解预测-校正扩散采样器的失配问题,显著提升无条件与条件采样的质量和效率。

摘要:

扩散概率模型(DPMs)在视觉合成中表现出卓越的性能,但由于采样过程中需要多次评估,计算成本较高。最近的预测-校正扩散采样器显著减少了所需的函数评估次数(NFE),但本质上受到由额外校正步骤引起的失配问题的影响,尤其是在无分类器引导尺度(CFG)较大时。本文提出了一种新的快速DPM采样器,称为DC-Solver,它利用动态补偿(DC)来缓解预测-校正采样器的失配问题。动态补偿由补偿比率控制,这些比率对采样步数具有自适应性,并且可以通过将采样轨迹推向真实轨迹,仅在10个数据点上进行优化。我们进一步提出了一种级联多项式回归(CPR),它可以即时预测未见采样配置下的补偿比率。此外,我们发现所提出的动态补偿还可以作为即插即用模块,提升仅预测采样器的性能。在无条件采样和条件采样上的大量实验表明,我们的DC-Solver能够在分辨率高达1024×1024的不同DPM上一致地提升采样质量,优于先前的方法。值得注意的是,我们在无条件FFHQ上取得了10.38 FID(NFE=5),在Stable-Diffusion-2.1上取得了0.394 MSE(NFE=5,CFG=7.5)。代码可在https://github.com/wl-zhao/DC-Solver获取。

英文摘要:

Diffusion probabilistic models (DPMs) have shown remarkable performance in visual synthesis but are computationally expensive due to the need for multiple evaluations during the sampling. Recent predictor-corrector diffusion samplers have significantly reduced the required number of function evaluations (NFE), but inherently suffer from a misalignment issue caused by the extra corrector step, especially with a large classifier-free guidance scale (CFG). In this paper, we introduce a new fast DPM sampler called DC-Solver, which leverages dynamic compensation (DC) to mitigate the misalignment of the predictor-corrector samplers. The dynamic compensation is controlled by compensation ratios that are adaptive to the sampling steps and can be optimized on only 10 datapoints by pushing the sampling trajectory toward a ground truth trajectory. We further propose a cascade polynomial regression (CPR) which can instantly predict the compensation ratios on unseen sampling configurations. Additionally, we find that the proposed dynamic compensation can also serve as a plug-and-play module to boost the performance of predictor-only samplers. Extensive experiments on both unconditional sampling and conditional sampling demonstrate that our DC-Solver can consistently improve the sampling quality over previous methods on different DPMs with a wide range of resolutions up to 1024$\times$1024. Notably, we achieve 10.38 FID (NFE=5) on unconditional FFHQ and 0.394 MSE (NFE=5, CFG=7.5) on Stable-Diffusion-2.1. Code is available at https://github.com/wl-zhao/DC-Solver

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑