Cross-Resolution Distribution Matching for Diffusion Distillation

TL;DR

This paper introduces Cross-Resolution Distribution Matching Distillation (RMD), a novel framework that accelerates diffusion-based image and video generation by bridging distribution gaps across resolutions, enabling high-fidelity, few-step inference.

Contribution

RMD proposes a new method for cross-resolution distribution matching using logSNR-based mapping and distribution alignment, significantly speeding up diffusion models without quality loss.

Findings

Achieves up to 33.4X speedup on SDXL

Preserves high visual fidelity during acceleration

Effective across various backbone architectures

Abstract

Diffusion distillation is central to accelerating image and video generation, yet existing methods are fundamentally limited by the denoising process, where step reduction has largely saturated. Partial timestep low-resolution generation can further accelerate inference, but it suffers noticeable quality degradation due to cross-resolution distribution gaps. We propose Cross-Resolution Distribution Matching Distillation (RMD), a novel distillation framework that bridges cross-resolution distribution gaps for high-fidelity, few-step multi-resolution cascaded inference. Specifically, RMD divides the timestep intervals for each resolution using logarithmic signal-to-noise ratio (logSNR) curves, and introduces logSNR-based mapping to compensate for resolution-induced shifts. Distribution matching is conducted along resolution trajectories to reduce the gap between low-resolution generator distributions and the teacher's high-resolution distribution. In addition, a predicted-noise re-injection mechanism is incorporated during upsampling to stabilize training and improve synthesis quality. Quantitative and qualitative results show that RMD preserves high-fidelity generation while accelerating inference across various backbones. Notably, RMD achieves up to 33.4X speedup on SDXL and 25.6X on Wan2.1-14B, while preserving high visual fidelity.