Paper page - One Editor, Many Edits: A Unified Training-Free Framework for Diverse Video Editing

Source: https://huggingface.co/papers/2609.04190

Abstract

EditVid is a unified training-free video editing framework that uses sparse causal memory, token injection, and soft latent blending to support both instruction-guided and subject-guided edits with high fidelity.

Video editing spans diverse editing paradigms, yet achieving high-quality instruction-guided andsubject-guided editingwithin a single unified framework remains challenging. We introduce EditVid, a training-free framework combiningsparse causal memoryfor local coherence,correspondence-based post-attention token injectionfor long-range identity preservation, andsoft latent blendingfor edit locality. The same framework supports instruction-guided and reference-guided edits, including style transfer, attribute modification, object insertion, part-level editing, and subject replacement. On FiVE, EditVid achieves 78.16FiVE-Acc, compared with 58.95 for the strongest evaluated training-free baseline, while obtaining competitive results on IVEBench. A user study further shows a 51.8\% overall preference for EditVid over 7 competing methods.

View arXiv pageView PDFProject pageGitHub0Add to collection

Get this paper in your agent:

hf papers read 2609\.04190

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2609.04190 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2609.04190 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2609.04190 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.