NaRCan: Natural Refined Canonical Image with Integration of Diffusion Prior for Video Editing
This work addresses the challenge of producing coherent and high-quality edited video sequences, which is an incremental improvement over current canonical-based methods in video editing.
The authors tackled the problem of generating high-quality natural canonical images for video editing by integrating a hybrid deformation field and diffusion prior, achieving a 14x training speedup and outperforming existing methods in video editing tasks.
We propose a video editing framework, NaRCan, which integrates a hybrid deformation field and diffusion prior to generate high-quality natural canonical images to represent the input video. Our approach utilizes homography to model global motion and employs multi-layer perceptrons (MLPs) to capture local residual deformations, enhancing the model's ability to handle complex video dynamics. By introducing a diffusion prior from the early stages of training, our model ensures that the generated images retain a high-quality natural appearance, making the produced canonical images suitable for various downstream tasks in video editing, a capability not achieved by current canonical-based methods. Furthermore, we incorporate low-rank adaptation (LoRA) fine-tuning and introduce a noise and diffusion prior update scheduling technique that accelerates the training process by 14 times. Extensive experimental results show that our method outperforms existing approaches in various video editing tasks and produces coherent and high-quality edited video sequences. See our project page for video results at https://koi953215.github.io/NaRCan_page/.