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

TL;DR

DFVEdit is a fast, zero-shot video editing method for Video Diffusion Transformers that operates on latents via flow transformation, significantly reducing computational costs while maintaining high editing quality.

Contribution

It introduces the Conditional Delta Flow Vector (CDFV), enabling efficient, unbiased flow-based editing without attention modification or fine-tuning.

Findings

At least 20x inference speed-up compared to existing methods.

85% memory reduction during editing.

State-of-the-art performance on structural fidelity and consistency.

Abstract

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.