DFVEdit: Conditional Delta Flow Vector for Zero-shot Video Editing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cai, Lingling, Zhao, Kang, Yuan, Hangjie, Wang, Xiang, Zhang, Yingya, Huang, Kejie
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911025683496960
author Cai, Lingling
Zhao, Kang
Yuan, Hangjie
Wang, Xiang
Zhang, Yingya
Huang, Kejie
author_facet Cai, Lingling
Zhao, Kang
Yuan, Hangjie
Wang, Xiang
Zhang, Yingya
Huang, Kejie
contents The advent of Video Diffusion Transformers (Video DiTs) marks a milestone in video generation. However, directly applying existing video editing methods to Video DiTs often incurs substantial computational overhead, due to resource-intensive attention modification or finetuning. To alleviate this problem, we present DFVEdit, an efficient zero-shot video editing method tailored for Video DiTs. DFVEdit eliminates the need for both attention modification and fine-tuning by directly operating on clean latents via flow transformation. To be more specific, we observe that editing and sampling can be unified under the continuous flow perspective. Building upon this foundation, we propose the Conditional Delta Flow Vector (CDFV) -- a theoretically unbiased estimation of DFV -- and integrate Implicit Cross Attention (ICA) guidance as well as Embedding Reinforcement (ER) to further enhance editing quality. DFVEdit excels in practical efficiency, offering at least 20x inference speed-up and 85% memory reduction on Video DiTs compared to attention-engineering-based editing methods. Extensive quantitative and qualitative experiments demonstrate that DFVEdit can be seamlessly applied to popular Video DiTs (e.g., CogVideoX and Wan2.1), attaining state-of-the-art performance on structural fidelity, spatial-temporal consistency, and editing quality.
format Preprint
id arxiv_https___arxiv_org_abs_2506_20967
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DFVEdit: Conditional Delta Flow Vector for Zero-shot Video Editing
Cai, Lingling
Zhao, Kang
Yuan, Hangjie
Wang, Xiang
Zhang, Yingya
Huang, Kejie
Computer Vision and Pattern Recognition
Artificial Intelligence
The advent of Video Diffusion Transformers (Video DiTs) marks a milestone in video generation. However, directly applying existing video editing methods to Video DiTs often incurs substantial computational overhead, due to resource-intensive attention modification or finetuning. To alleviate this problem, we present DFVEdit, an efficient zero-shot video editing method tailored for Video DiTs. DFVEdit eliminates the need for both attention modification and fine-tuning by directly operating on clean latents via flow transformation. To be more specific, we observe that editing and sampling can be unified under the continuous flow perspective. Building upon this foundation, we propose the Conditional Delta Flow Vector (CDFV) -- a theoretically unbiased estimation of DFV -- and integrate Implicit Cross Attention (ICA) guidance as well as Embedding Reinforcement (ER) to further enhance editing quality. DFVEdit excels in practical efficiency, offering at least 20x inference speed-up and 85% memory reduction on Video DiTs compared to attention-engineering-based editing methods. Extensive quantitative and qualitative experiments demonstrate that DFVEdit can be seamlessly applied to popular Video DiTs (e.g., CogVideoX and Wan2.1), attaining state-of-the-art performance on structural fidelity, spatial-temporal consistency, and editing quality.
title DFVEdit: Conditional Delta Flow Vector for Zero-shot Video Editing
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2506.20967