NOVA: Sparse Control, Dense Synthesis for Pair-Free Video Editing

TL;DR

NOVA is a novel unpaired video editing framework that combines sparse semantic guidance with dense motion and texture synthesis, using a degradation-based training strategy to ensure high fidelity and temporal consistency without paired data.

Contribution

The paper introduces NOVA, a new framework for pair-free video editing that leverages sparse user guidance and dense synthesis, along with a degradation-simulation training method.

Findings

Outperforms existing methods in edit fidelity

Maintains high motion preservation

Ensures temporal coherence

Abstract

Recent video editing models have achieved impressive results, but most still require large-scale paired datasets. Collecting such naturally aligned pairs at scale remains highly challenging and constitutes a critical bottleneck, especially for local video editing data. Existing workarounds transfer image editing to video through global motion control for pair-free video editing, but such designs struggle with background and temporal consistency. In this paper, we propose NOVA: Sparse Control \& Dense Synthesis, a new framework for unpaired video editing. Specifically, the sparse branch provides semantic guidance through user-edited keyframes distributed across the video, and the dense branch continuously incorporates motion and texture information from the original video to maintain high fidelity and coherence. Moreover, we introduce a degradation-simulation training strategy that enables the model to learn motion reconstruction and temporal consistency by training on artificially degraded videos, thus eliminating the need for paired data. Our extensive experiments demonstrate that NOVA outperforms existing approaches in edit fidelity, motion preservation, and temporal coherence.