TokenFlow: Consistent Diffusion Features for Consistent Video Editing (ICLR 2024)

TokenFlow is a framework that enables consistent video editing, using a pre-trained text-to-image diffusion model, without any further training or finetuning.

The generative AI revolution has been recently expanded to videos. Nevertheless, current state-of-the-art video models are still lagging behind image models in terms of visual quality and user control over the generated content. In this work, we present a framework that harnesses the power of a text-to-image diffusion model for the task of text-driven video editing. Specifically, given a source video and a target text-prompt, our method generates a high-quality video that adheres to the target text, while preserving the spatial layout and dynamics of the input video. Our method is based on our key observation that consistency in the edited video can be obtained by enforcing consistency in the diffusion feature space. We achieve this by explicitly propagating diffusion features based on inter-frame correspondences, readily available in the model. Thus, our framework does not require any training or fine-tuning, and can work in conjunction with any off-the-shelf text-to-image editing method. We demonstrate state-of-the-art editing results on a variety of real-world videos.

Citation

@article{tokenflow2023,
        title = {TokenFlow: Consistent Diffusion Features for Consistent Video Editing},
        author = {Geyer, Michal and Bar-Tal, Omer and Bagon, Shai and Dekel, Tali},
        journal={arXiv preprint arxiv:2307.10373},
        year={2023}
        }