Reverse Transition Kernel: A Flexible Framework to Accelerate Diffusion Inference

• URL: http://arxiv.org/abs/2405.16387v1
• Date: Sun, 26 May 2024 00:26:57 GMT
• Title: Reverse Transition Kernel: A Flexible Framework to Accelerate Diffusion Inference
• Authors: Xunpeng Huang, Difan Zou, Hanze Dong, Yi Zhang, Yi-An Ma, Tong Zhang,
• Abstract summary: DDPM uses a Gaussian approximation for the RTK, resulting in low per-subproblem complexity but requiring a large number of segments. We develop a general RTK framework that enables a more balanced subproblem decomposition, resulting in $tilde O(1)$ subproblems. We then propose leveraging two fast sampling algorithms, the Metropolis-Adjusted Langevin Algorithm (MALA) and Underdamped Langevin Dynamics (ULD) for solving these strongly log-concave subproblems.
• Score: 29.589117191576875
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: To generate data from trained diffusion models, most inference algorithms, such as DDPM, DDIM, and other variants, rely on discretizing the reverse SDEs or their equivalent ODEs. In this paper, we view such approaches as decomposing the entire denoising diffusion process into several segments, each corresponding to a reverse transition kernel (RTK) sampling subproblem. Specifically, DDPM uses a Gaussian approximation for the RTK, resulting in low per-subproblem complexity but requiring a large number of segments (i.e., subproblems), which is conjectured to be inefficient. To address this, we develop a general RTK framework that enables a more balanced subproblem decomposition, resulting in $\tilde O(1)$ subproblems, each with strongly log-concave targets. We then propose leveraging two fast sampling algorithms, the Metropolis-Adjusted Langevin Algorithm (MALA) and Underdamped Langevin Dynamics (ULD), for solving these strongly log-concave subproblems. This gives rise to the RTK-MALA and RTK-ULD algorithms for diffusion inference. In theory, we further develop the convergence guarantees for RTK-MALA and RTK-ULD in total variation (TV) distance: RTK-ULD can achieve $\epsilon$ target error within $\tilde{\mathcal O}(d^{1/2}\epsilon^{-1})$ under mild conditions, and RTK-MALA enjoys a $\mathcal{O}(d^{2}\log(d/\epsilon))$ convergence rate under slightly stricter conditions. These theoretical results surpass the state-of-the-art convergence rates for diffusion inference and are well supported by numerical experiments.

Related papers

• Nested Stochastic Gradient Descent for (Generalized) Sinkhorn Distance-Regularized Distributionally Robust Optimization [4.989068568135242]
  Distributionally robust shift optimization (DRO) is a powerful technique to robust models against data distribution. This paper aims to solve regularized non DRO problems, where the uncertainty is modeled by a so-called generalized approximation function.
  arXiv  Detail & Related papers  (2025-03-29T01:01:02Z)
• Latent Schrodinger Bridge: Prompting Latent Diffusion for Fast Unpaired Image-to-Image Translation [58.19676004192321]
  Diffusion models (DMs), which enable both image generation from noise and inversion from data, have inspired powerful unpaired image-to-image (I2I) translation algorithms. We tackle this problem with Schrodinger Bridges (SBs), which are differential equations (SDEs) between distributions with minimal transport cost. Inspired by this observation, we propose Latent Schrodinger Bridges (LSBs) that approximate the SB ODE via pre-trained Stable Diffusion. We demonstrate that our algorithm successfully conduct competitive I2I translation in unsupervised setting with only a fraction of cost required by previous DM-
  arXiv  Detail & Related papers  (2024-11-22T11:24:14Z)
• Faster Sampling via Stochastic Gradient Proximal Sampler [28.422547264326468]
  Proximal samplers (SPS) for sampling from non-log-concave distributions are studied. We show that the convergence to the target distribution can be guaranteed as long as the algorithm trajectory is bounded. We provide two implementable variants based on Langevin dynamics (SGLD) and Langevin-MALA, giving rise to SPS-SGLD and SPS-MALA.
  arXiv  Detail & Related papers  (2024-05-27T00:53:18Z)
• An Improved Analysis of Langevin Algorithms with Prior Diffusion for Non-Log-Concave Sampling [27.882407333690267]
  We show that modified Langevin algorithm with prior diffusion can converge dimension independently for strongly log-concave target distributions. We also prove that the modified Langevin algorithm can also obtain the dimension-independent convergence of KL divergence with different step size schedules.
  arXiv  Detail & Related papers  (2024-03-10T11:50:34Z)
• Faster Sampling without Isoperimetry via Diffusion-based Monte Carlo [30.4930148381328]
  Diffusion-based Monte Carlo (DMC) is a method to sample from a general target distribution beyond the isoperimetric condition. DMC encountered high gradient complexity, resulting in an exponential dependency on the error tolerance $epsilon$ of the obtained samples. We propose RS-DMC, based on a novel recursion-based score estimation method. Our algorithm is provably much faster than the popular Langevin-based algorithms.
  arXiv  Detail & Related papers  (2024-01-12T02:33:57Z)
• Gaussian Mixture Solvers for Diffusion Models [84.83349474361204]
  We introduce a novel class of SDE-based solvers called GMS for diffusion models. Our solver outperforms numerous SDE-based solvers in terms of sample quality in image generation and stroke-based synthesis.
  arXiv  Detail & Related papers  (2023-11-02T02:05:38Z)
• Stochastic Optimization for Non-convex Problem with Inexact Hessian Matrix, Gradient, and Function [99.31457740916815]
  Trust-region (TR) and adaptive regularization using cubics have proven to have some very appealing theoretical properties. We show that TR and ARC methods can simultaneously provide inexact computations of the Hessian, gradient, and function values.
  arXiv  Detail & Related papers  (2023-10-18T10:29:58Z)
• Decomposed Diffusion Sampler for Accelerating Large-Scale Inverse Problems [64.29491112653905]
  We propose a novel and efficient diffusion sampling strategy that synergistically combines the diffusion sampling and Krylov subspace methods. Specifically, we prove that if tangent space at a denoised sample by Tweedie's formula forms a Krylov subspace, then the CG with the denoised data ensures the data consistency update to remain in the tangent space. Our proposed method achieves more than 80 times faster inference time than the previous state-of-the-art method.
  arXiv  Detail & Related papers  (2023-03-10T07:42:49Z)
• Faster high-accuracy log-concave sampling via algorithmic warm starts [6.117084972237769]
  In practice, high-accuracy samplers such as the classical Metropolis-adjusted Langevin algorithm (MALA) remain the de facto gold standard. We improve the dimension dependence of this sampling problem to $tildeO(d1/2)$, whereas the previous best result for MALA was $tildeO(d)$. Our main technical contribution settles this problem by establishing the first $tildeO(d1/2)$ R'enyi mixing rates for the discretized underdamped Langevin diffusion.
  arXiv  Detail & Related papers  (2023-02-20T19:27:21Z)
• DR-DSGD: A Distributionally Robust Decentralized Learning Algorithm over Graphs [54.08445874064361]
  We propose to solve a regularized distributionally robust learning problem in the decentralized setting. By adding a Kullback-Liebler regularization function to the robust min-max optimization problem, the learning problem can be reduced to a modified robust problem. We show that our proposed algorithm can improve the worst distribution test accuracy by up to $10%$.
  arXiv  Detail & Related papers  (2022-08-29T18:01:42Z)
• Improved Convergence Rate of Stochastic Gradient Langevin Dynamics with Variance Reduction and its Application to Optimization [50.83356836818667]
  gradient Langevin Dynamics is one of the most fundamental algorithms to solve non-eps optimization problems. In this paper, we show two variants of this kind, namely the Variance Reduced Langevin Dynamics and the Recursive Gradient Langevin Dynamics.
  arXiv  Detail & Related papers  (2022-03-30T11:39:00Z)
• Distributed Random Reshuffling over Networks [7.013052033764372]
  A distributed resh-upr (D-RR) algorithm is proposed to solve the problem of convex and smooth objective functions. In particular, for smooth convex objective functions, D-RR achieves D-T convergence rate (where $T counts epoch number) in terms of distance between the global drives.
  arXiv  Detail & Related papers  (2021-12-31T03:59:37Z)
• Distributed stochastic optimization with large delays [59.95552973784946]
  One of the most widely used methods for solving large-scale optimization problems is distributed asynchronous gradient descent (DASGD) We show that DASGD converges to a global optimal implementation model under same delay assumptions.
  arXiv  Detail & Related papers  (2021-07-06T21:59:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.