Transform image editors into video editors using keyframe anchors
Transforming Image Editors into Video Editors
Computer Vision and Pattern Recognition
Summary
Editing videos can be much harder and more expensive than editing single images. The authors found that you can make video editing easier by first editing key pictures in the video and then filling in the rest using a special method. They built a system that uses a strong image editor to change important frames and then creates the final video by adding consistent motion over time. This approach works well and shows future video editing can improve by building on better image editing tools.
What this means in practice
- •For video content creators: Create edited videos by modifying only key frames with image editing tools and automatically generate the rest of the video consistently.
- •For visual effects teams: Produce high-quality video edits using existing image editing software combined with motion-guided video generation to save time and cost.
Authors
Feng Wang, Zijie Li, Ceyuan Yang, Alan Yuille, Peng Wang
Abstract
Recent image editing systems have achieved impressive semantic understanding, visual fidelity, and instruction-following ability, while video editing remains substantially more difficult and costly. In this paper, we present a simple alternative to end-to-end video editing: instead of training a monolithic video editor, we transform a strong image editor into a video editor through anchor-based generation. Our key insight is that video editing can be decomposed into two subproblems: editing a sparse set of keyframes and propagating those edits across time. Based on this observation, we propose Anchor-based Video Editing (AVE), a two-stage framework in which a powerful image editor first performs composed editing on selected keyframes, and a motion-guided image-to-video diffusion model then generates the final video by treating the edited keyframes as fixed anchors. This design directly inherits the strengths of modern image editors while avoiding expensive end-to-end video editing training. Experiments on IVEBench and VIE-Bench show that AVE achieves strong performance in instruction following, temporal consistency, and content fidelity. Further ablations reveal that final video editing quality is strongly correlated with the quality of the image editor, suggesting that future progress in video editing may come from stronger image editing foundations and lightweight transfer to video. Code is available at https://github.com/wangf3014/AVE.