Continuous-Time Distribution Matching for Few-Step Diffusion Distillation

TL;DR

This paper introduces Continuous-Time Distribution Matching (CDM), a novel approach that enhances diffusion model distillation by enforcing distribution matching along continuous trajectories, leading to improved visual fidelity in few-step image generation.

Contribution

It pioneers the migration of Distribution Matching Distillation from discrete to continuous optimization, enabling arbitrary point enforcement and active off-trajectory matching.

Findings

CDM achieves highly competitive visual fidelity in few-step image generation.

It outperforms traditional discrete DMD in preserving fine visual details.

The approach reduces reliance on complex auxiliary modules.

Abstract

Step distillation has become a leading technique for accelerating diffusion models, among which Distribution Matching Distillation (DMD) and Consistency Distillation are two representative paradigms. While consistency methods enforce self-consistency along the full PF-ODE trajectory to steer it toward the clean data manifold, vanilla DMD relies on sparse supervision at a few predefined discrete timesteps. This restricted discrete-time formulation and mode-seeking nature of the reverse KL divergence tends to exhibit visual artifacts and over-smoothed outputs, often necessitating complex auxiliary modules -- such as GANs or reward models -- to restore visual fidelity. In this work, we introduce Continuous-Time Distribution Matching (CDM), migrating the DMD framework from discrete anchoring to continuous optimization for the first time. CDM achieves this through two continuous-time designs. First, we replace the fixed discrete schedule with a dynamic continuous schedule of random length, so that distribution matching is enforced at arbitrary points along sampling trajectories rather than only at a few fixed anchors. Second, we propose a continuous-time alignment objective that performs active off-trajectory matching on latents extrapolated via the student's velocity field, improving generalization and preserving fine visual details. Extensive experiments on different architectures, including SD3-Medium and Longcat-Image, demonstrate that CDM provides highly competitive visual fidelity for few-step image generation without relying on complex auxiliary objectives. Code is available at https://github.com/byliutao/cdm.