Cross-Frame Representation Alignment for Fine-Tuning Video Diffusion Models

TL;DR

This paper introduces CREPA, a novel regularization technique for fine-tuning Video Diffusion Models that enhances both visual quality and temporal semantic consistency across frames.

Contribution

It adapts the Representation Alignment method for VDMs and proposes CREPA to improve cross-frame semantic coherence during fine-tuning.

Findings

CREPA improves visual fidelity in fine-tuned VDMs.

CREPA enhances cross-frame semantic consistency.

Empirical results on large-scale VDMs validate CREPA's effectiveness.

Abstract

Fine-tuning Video Diffusion Models (VDMs) at the user level to generate videos that reflect specific attributes of training data presents notable challenges, yet remains underexplored despite its practical importance. Meanwhile, recent work such as Representation Alignment (REPA) has shown promise in improving the convergence and quality of DiT-based image diffusion models by aligning, or assimilating, its internal hidden states with external pretrained visual features, suggesting its potential for VDM fine-tuning. In this work, we first propose a straightforward adaptation of REPA for VDMs and empirically show that, while effective for convergence, it is suboptimal in preserving semantic consistency across frames. To address this limitation, we introduce Cross-frame Representation Alignment (CREPA), a novel regularization technique that aligns hidden states of a frame with external features from neighboring frames. Empirical evaluations on large-scale VDMs, including CogVideoX-5B and Hunyuan Video, demonstrate that CREPA improves both visual fidelity and cross-frame semantic coherence when fine-tuned with parameter-efficient methods such as LoRA. We further validate CREPA across diverse datasets with varying attributes, confirming its broad applicability.