Streaming Video Diffusion: Online Video Editing with Diffusion Models

TL;DR

This paper introduces Streaming Video Diffusion, a real-time online video editing method that maintains temporal consistency in streaming videos using a diffusion model with temporal recurrence, suitable for live applications.

Contribution

The paper proposes SVDiff, a novel diffusion-based framework for online video editing that handles streaming frames with temporal coherence and zero-shot capabilities.

Findings

Achieves 15.2 FPS inference speed at 512x512 resolution.

Effectively edits long, high-quality videos with temporal consistency.

Supports a broad range of videos with a single model.

Abstract

We present a novel task called online video editing, which is designed to edit \textbf{streaming} frames while maintaining temporal consistency. Unlike existing offline video editing assuming all frames are pre-established and accessible, online video editing is tailored to real-life applications such as live streaming and online chat, requiring (1) fast continual step inference, (2) long-term temporal modeling, and (3) zero-shot video editing capability. To solve these issues, we propose Streaming Video Diffusion (SVDiff), which incorporates the compact spatial-aware temporal recurrence into off-the-shelf Stable Diffusion and is trained with the segment-level scheme on large-scale long videos. This simple yet effective setup allows us to obtain a single model that is capable of executing a broad range of videos and editing each streaming frame with temporal coherence. Our experiments indicate that our model can edit long, high-quality videos with remarkable results, achieving a real-time inference speed of 15.2 FPS at a resolution of 512x512.