InstructX: Towards Unified Visual Editing with MLLM Guidance
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866908585228763136 |
|---|---|
| author | Mou, Chong Sun, Qichao Wu, Yanze Zhang, Pengze Li, Xinghui Ye, Fulong Zhao, Songtao He, Qian |
| author_facet | Mou, Chong Sun, Qichao Wu, Yanze Zhang, Pengze Li, Xinghui Ye, Fulong Zhao, Songtao He, Qian |
| contents | With recent advances in Multimodal Large Language Models (MLLMs) showing strong visual understanding and reasoning, interest is growing in using them to improve the editing performance of diffusion models. Despite rapid progress, most studies lack an in-depth analysis of MLLM design choices. Moreover, the integration of MLLMs and diffusion models remains an open challenge in some difficult tasks, such as video editing. In this paper, we present InstructX, a unified framework for image and video editing. Specifically, we conduct a comprehensive study on integrating MLLMs and diffusion models for instruction-driven editing across diverse tasks. Building on this study, we analyze the cooperation and distinction between images and videos in unified modeling. (1) We show that training on image data can lead to emergent video editing capabilities without explicit supervision, thereby alleviating the constraints imposed by scarce video training data. (2) By incorporating modality-specific MLLM features, our approach effectively unifies image and video editing tasks within a single model. Extensive experiments demonstrate that our method can handle a broad range of image and video editing tasks and achieves state-of-the-art performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_08485 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | InstructX: Towards Unified Visual Editing with MLLM Guidance Mou, Chong Sun, Qichao Wu, Yanze Zhang, Pengze Li, Xinghui Ye, Fulong Zhao, Songtao He, Qian Computer Vision and Pattern Recognition With recent advances in Multimodal Large Language Models (MLLMs) showing strong visual understanding and reasoning, interest is growing in using them to improve the editing performance of diffusion models. Despite rapid progress, most studies lack an in-depth analysis of MLLM design choices. Moreover, the integration of MLLMs and diffusion models remains an open challenge in some difficult tasks, such as video editing. In this paper, we present InstructX, a unified framework for image and video editing. Specifically, we conduct a comprehensive study on integrating MLLMs and diffusion models for instruction-driven editing across diverse tasks. Building on this study, we analyze the cooperation and distinction between images and videos in unified modeling. (1) We show that training on image data can lead to emergent video editing capabilities without explicit supervision, thereby alleviating the constraints imposed by scarce video training data. (2) By incorporating modality-specific MLLM features, our approach effectively unifies image and video editing tasks within a single model. Extensive experiments demonstrate that our method can handle a broad range of image and video editing tasks and achieves state-of-the-art performance. |
| title | InstructX: Towards Unified Visual Editing with MLLM Guidance |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.08485 |