GRNEdit: Efficient General Video Editing from a New Binary-Evidence Perspective in Generative Refinement Networks

Hugging Face Daily Papers Papers
video-editing ai-research generative-models efficient-ai binary-evidence fine-tuning

Summary

GRNEdit is a lightweight two-stage framework for efficient general video editing that uses binary semantic decisions and source evidence to achieve high performance with fewer parameters than larger models.

Instruction-based general video editing seeks to unify diverse editing operations within a single, intuitive interface. Existing approaches often rely on resource-intensive conditioning, using either heavyweight branches or costly source concatenation. Is there any efficient way to model editing intent? Thus, we introduce GRNEdit, a lightweight two-stage framework. GRN inspires our approach by encoding visual semantics through combinations of bits. Through task-specific fine-tuning, we take this representation further and recast editing semantics as local retain-or-flip decisions over individual bits. Source information is consequently modeled as coordinate-wise evidence supporting the observed binary states, while the GRN backbone remains responsible for resolving their global composition into coherent generative semantics. In Stage I, a compact encoder translates discrete source codes into continuous evidence signals, which GRN assimilates throughout binary refinement. Inspired by null-prompt training for classifier-free guidance, we further assign the null condition an editing-specific meaning: an empty instruction denotes no edit and is supervised through source reconstruction. This identity pathway not only implicitly strengthens evidence utilization and content preservation in Stage I, but also produces a source-preserving state in the same representation space as the edited state. Stage II can therefore directly compare each edited state with its source-preserving counterpart and use their discrepancy to revise unresolved target-bit decisions. Trained on only 0.6M pairs with less than 3\% conditioning parameters, GRNEdit-2B and GRNEdit-8B achieve scores of 4.03 and 4.18 on OpenVE-Bench. The 2B model outperforms multiple 14B open-source editors, while the 8B model performs on par with leading open-source editors.
Original Article
View Cached Full Text

Cached at: 08/18/26, 11:51 AM

Paper page - GRNEdit: Efficient General Video Editing from a New Binary-Evidence Perspective in Generative Refinement Networks

Source: https://huggingface.co/papers/2608.16328 Published on Aug 17

·

Submitted byhttps://huggingface.co/HigherHu

Higheron Aug 18

Abstract

GRNEdit is a lightweight two-stage framework that models video editing intent via binary semantic decisions and source evidence, achieving strong results with minimal parameters.

Instruction-based general video editing seeks to unify diverse editing operations within a single, intuitive interface. Existing approaches often rely on resource-intensive conditioning, using either heavyweight branches or costly source concatenation. Is there any efficient way to model editing intent? Thus, we introduceGRNEdit, a lightweight two-stage framework. GRN inspires our approach by encoding visual semantics through combinations of bits. Through task-specific fine-tuning, we take this representation further and recast editing semantics as local retain-or-flip decisions over individual bits. Source information is consequently modeled ascoordinate-wise evidencesupporting the observed binary states, while the GRN backbone remains responsible for resolving their global composition into coherent generative semantics. In Stage I, a compact encoder translates discrete source codes into continuous evidence signals, which GRN assimilates throughoutbinary refinement. Inspired bynull-prompt trainingforclassifier-free guidance, we further assign the null condition an editing-specific meaning: an empty instruction denotes no edit and is supervised through source reconstruction. Thisidentity pathwaynot only implicitly strengthens evidence utilization and content preservation in Stage I, but also produces asource-preserving statein the same representation space as the edited state. Stage II can therefore directly compare each edited state with its source-preserving counterpart and use their discrepancy to revise unresolved target-bit decisions. Trained on only 0.6M pairs with less than 3\% conditioning parameters,GRNEdit-2B andGRNEdit-8B achieve scores of 4.03 and 4.18 onOpenVE-Bench. The 2B model outperforms multiple 14B open-source editors, while the 8B model performs on par with leading open-source editors.

View arXiv pageView PDFProject pageGitHub4Add to collection

Get this paper in your agent:

hf papers read 2608\.16328

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/2608.16328 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2608.16328 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2608.16328 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.

Similar Articles

Streaming Video Editing with Easy Adaptation

Hugging Face Daily Papers

This paper introduces SVEET, a framework for high-quality streaming video editing that leverages a pretrained video diffusion model to enable auto-regressive editing with real-time performance on a single GPU.