OSVE: One Step Video Editing with One Step Diffusion Models

2026-07-22

AI Summary

This work addresses the high computational cost and low efficiency of existing text-guided video editing methods, which rely on multi-step sampling and inversion. It presents the first successful extension of single-step text-to-image diffusion models to video editing by introducing a learnable noise encoder that enables noise prediction within a single forward pass. To ensure temporal consistency and high editing fidelity, the approach incorporates a Structure-Aware Editing (SAE) loss and a Unified Frame Editing (UFE) mechanism, leveraging a sliding window strategy with anchor frames. The proposed method achieves comparable or superior editing quality relative to state-of-the-art multi-step techniques while accelerating inference by 155–171 times, substantially advancing the practicality of real-time video editing.

Abstract

Text-guided video editing with diffusion models is impractically slow, hindered by costly multi-step sampling and inversion. We present OSVE, the first framework to successfully adapt one-step Text-to-Image (T2I) models for high-quality video editing, addressing the core challenges of inversion, editability, and temporal consistency. To bypass slow iterative inversion, we train a learnable encoder that predicts the initial noise for each frame in a single forward pass. This encoder is trained with a novel Structure-Aware Editing (SAE) loss on a curated dataset of structurally-aligned image pairs, teaching it to preserve the source video's geometry during edits. For temporal coherence, we introduce Unified-Frame Editing (UFE), a technique that concatenates frame latents to facilitate cross-frame attention in a single generation step. Furthermore, for long videos, a sliding-window strategy with an anchor frame maintains global consistency. Our extensive experiments demonstrate that OSVE achieves editing quality comparable or superior to state-of-the-art multi-step methods, while operating approximately 155--171 times faster. This breakthrough paves the way for practical, real-time video editing applications. Code is available at https://github.com/KU-VGI/OSVE.

Problem

• text-guided video editing
• diffusion models
• temporal consistency
• inversion
• computational efficiency

Innovation

• one-step diffusion
• video editing
• temporal consistency
• structure-aware editing
• learnable inversion