Flow-Guided Diffusion for Video Inpainting

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Gu, Bohai, Yu, Yongsheng, Fan, Heng, Zhang, Libo
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909464032968704
author Gu, Bohai
Yu, Yongsheng
Fan, Heng
Zhang, Libo
author_facet Gu, Bohai
Yu, Yongsheng
Fan, Heng
Zhang, Libo
contents Video inpainting has been challenged by complex scenarios like large movements and low-light conditions. Current methods, including emerging diffusion models, face limitations in quality and efficiency. This paper introduces the Flow-Guided Diffusion model for Video Inpainting (FGDVI), a novel approach that significantly enhances temporal consistency and inpainting quality via reusing an off-the-shelf image generation diffusion model. We employ optical flow for precise one-step latent propagation and introduces a model-agnostic flow-guided latent interpolation technique. This technique expedites denoising, seamlessly integrating with any Video Diffusion Model (VDM) without additional training. Our FGDVI demonstrates a remarkable 10% improvement in flow warping error E_warp over existing state-of-the-art methods. Our comprehensive experiments validate superior performance of FGDVI, offering a promising direction for advanced video inpainting. The code and detailed results will be publicly available in https://github.com/NevSNev/FGDVI.
format Preprint
id arxiv_https___arxiv_org_abs_2311_15368
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Flow-Guided Diffusion for Video Inpainting
Gu, Bohai
Yu, Yongsheng
Fan, Heng
Zhang, Libo
Computer Vision and Pattern Recognition
Video inpainting has been challenged by complex scenarios like large movements and low-light conditions. Current methods, including emerging diffusion models, face limitations in quality and efficiency. This paper introduces the Flow-Guided Diffusion model for Video Inpainting (FGDVI), a novel approach that significantly enhances temporal consistency and inpainting quality via reusing an off-the-shelf image generation diffusion model. We employ optical flow for precise one-step latent propagation and introduces a model-agnostic flow-guided latent interpolation technique. This technique expedites denoising, seamlessly integrating with any Video Diffusion Model (VDM) without additional training. Our FGDVI demonstrates a remarkable 10% improvement in flow warping error E_warp over existing state-of-the-art methods. Our comprehensive experiments validate superior performance of FGDVI, offering a promising direction for advanced video inpainting. The code and detailed results will be publicly available in https://github.com/NevSNev/FGDVI.
title Flow-Guided Diffusion for Video Inpainting
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.15368