Are Image-to-Video Models Good Zero-Shot Image Editors?

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Zhang, Zechuan, Chen, Zhenyuan, Yang, Zongxin, Yang, Yi
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909924070522880
author Zhang, Zechuan
Chen, Zhenyuan
Yang, Zongxin
Yang, Yi
author_facet Zhang, Zechuan
Chen, Zhenyuan
Yang, Zongxin
Yang, Yi
contents Large-scale video diffusion models show strong world simulation and temporal reasoning abilities, but their use as zero-shot image editors remains underexplored. We introduce IF-Edit, a tuning-free framework that repurposes pretrained image-to-video diffusion models for instruction-driven image editing. IF-Edit addresses three key challenges: prompt misalignment, redundant temporal latents, and blurry late-stage frames. It includes (1) a chain-of-thought prompt enhancement module that transforms static editing instructions into temporally grounded reasoning prompts; (2) a temporal latent dropout strategy that compresses frame latents after the expert-switch point, accelerating denoising while preserving semantic and temporal coherence; and (3) a self-consistent post-refinement step that sharpens late-stage frames using a short still-video trajectory. Experiments on four public benchmarks, covering non-rigid editing, physical and temporal reasoning, and general instruction edits, show that IF-Edit performs strongly on reasoning-centric tasks while remaining competitive on general-purpose edits. Our study provides a systematic view of video diffusion models as image editors and highlights a simple recipe for unified video-image generative reasoning.
format Preprint
id arxiv_https___arxiv_org_abs_2511_19435
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Are Image-to-Video Models Good Zero-Shot Image Editors?
Zhang, Zechuan
Chen, Zhenyuan
Yang, Zongxin
Yang, Yi
Computer Vision and Pattern Recognition
Large-scale video diffusion models show strong world simulation and temporal reasoning abilities, but their use as zero-shot image editors remains underexplored. We introduce IF-Edit, a tuning-free framework that repurposes pretrained image-to-video diffusion models for instruction-driven image editing. IF-Edit addresses three key challenges: prompt misalignment, redundant temporal latents, and blurry late-stage frames. It includes (1) a chain-of-thought prompt enhancement module that transforms static editing instructions into temporally grounded reasoning prompts; (2) a temporal latent dropout strategy that compresses frame latents after the expert-switch point, accelerating denoising while preserving semantic and temporal coherence; and (3) a self-consistent post-refinement step that sharpens late-stage frames using a short still-video trajectory. Experiments on four public benchmarks, covering non-rigid editing, physical and temporal reasoning, and general instruction edits, show that IF-Edit performs strongly on reasoning-centric tasks while remaining competitive on general-purpose edits. Our study provides a systematic view of video diffusion models as image editors and highlights a simple recipe for unified video-image generative reasoning.
title Are Image-to-Video Models Good Zero-Shot Image Editors?
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.19435