# MiVE: Multiscale Vision-language features for reference-guided video Editing

arXiv: https://arxiv.org/abs/2605.14664 · HTML: https://mlsanity.com/paper/2605.14664/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 26 references that are in the map.
- First posted on arXiv 14 May 2026.

## Builds on (most cited first, 18 of 18)

- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency](https://mlsanity.com/paper/2508.18265/) (2025)
- [Qwen-Image Technical Report](https://mlsanity.com/paper/2508.02324/) (2025)
- [FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space](https://mlsanity.com/paper/2506.15742/) (2025)
- [VACE: All-in-One Video Creation and Editing](https://mlsanity.com/paper/2503.07598/) (2025)
- [Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer](https://mlsanity.com/paper/2511.22699/) (2025)
- [MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe](https://mlsanity.com/paper/2509.18154/) (2025)
- [Wan-Animate: Unified Character Animation and Replacement with Holistic Replication](https://mlsanity.com/paper/2509.14055/) (2025)
- [Scaling Instruction-Based Video Editing with a High-Quality Synthetic Dataset](https://mlsanity.com/paper/2510.15742/) (2025)
- [Kling-Omni Technical Report](https://mlsanity.com/paper/2512.16776/) (2025)
- [OpenVE-3M: A Large-Scale High-Quality Dataset for Instruction-Guided Video Editing](https://mlsanity.com/paper/2512.07826/) (2025)
- [ROSE: Remove Objects with Side Effects in Videos](https://mlsanity.com/paper/2508.18633/) (2025)
- [In-Context Learning with Unpaired Clips for Instruction-based Video Editing](https://mlsanity.com/paper/2510.14648/) (2025)
- [BindWeave: Subject-Consistent Video Generation via Cross-Modal Integration](https://mlsanity.com/paper/2510.00438/) (2025)
- [Region-Constraint In-Context Generation for Instructional Video Editing](https://mlsanity.com/paper/2512.17650/) (2025)
- [VideoCoF: Unified Video Editing with Temporal Reasoner](https://mlsanity.com/paper/2512.07469/) (2025)
- [LoVoRA: Text-guided and Mask-free Video Object Removal and Addition with Learnable Object-aware Localization](https://mlsanity.com/paper/2512.02933/) (2025)

## Built on by (most cited first, 1 of 1)

- [TIDE: Task-Isolated Diffusion for Unified Video Editing and Generation](https://mlsanity.com/paper/2606.08260/) (2026)

## Abstract (arXiv)

Reference-guided video editing takes a source video, a text instruction, and a reference image as inputs, requiring the model to faithfully apply the instructed edits while preserving original motion and unedited content. Existing methods fall into two paradigms, each with inherent limitations: decoupled encoders suffer from modality gaps when processing instructions and visual content independently, while unified vision-language encoders lose fine-grained spatial details by relying solely on final-layer representations. We observe that VLM layers encode complementary information hierarchically -- early layers capture localized spatial details essential for precise editing, while deeper layers encode global semantics for instruction comprehension. Building on this insight, we present MiVE (Multiscale Vision-language features for reference-guided video Editing), a framework that repurposes VLMs as multiscale feature extractors. MiVE extracts hierarchical features from Qwen3-VL and integrates them into a unified self-attention Diffusion Transformer, eliminating the modality mismatch inherent in cross-attention designs. Experiments demonstrate that MiVE achieves state-of-the-art performance by ranking highest in human preference, outperforming both academic methods and commercial systems.