Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understanding
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_ | 1866918156277121024 |
|---|---|
| author | Xin, Yi Qin, Qi Luo, Siqi Zhu, Kaiwen Yan, Juncheng Tai, Yan Lei, Jiayi Cao, Yuewen Wang, Keqi Wang, Yibin Bai, Jinbin Yu, Qian Jiang, Dengyang Pu, Yuandong Chen, Haoxing Zhuo, Le He, Junjun Luo, Gen Li, Tianbin Hu, Ming Ye, Jin Ye, Shenglong Zhang, Bo Xu, Chang Wang, Wenhai Li, Hongsheng Zhai, Guangtao Xue, Tianfan Fu, Bin Liu, Xiaohong Qiao, Yu Liu, Yihao |
| author_facet | Xin, Yi Qin, Qi Luo, Siqi Zhu, Kaiwen Yan, Juncheng Tai, Yan Lei, Jiayi Cao, Yuewen Wang, Keqi Wang, Yibin Bai, Jinbin Yu, Qian Jiang, Dengyang Pu, Yuandong Chen, Haoxing Zhuo, Le He, Junjun Luo, Gen Li, Tianbin Hu, Ming Ye, Jin Ye, Shenglong Zhang, Bo Xu, Chang Wang, Wenhai Li, Hongsheng Zhai, Guangtao Xue, Tianfan Fu, Bin Liu, Xiaohong Qiao, Yu Liu, Yihao |
| contents | We introduce Lumina-DiMOO, an open-source foundational model for seamless multi-modal generation and understanding. Lumina-DiMOO sets itself apart from prior unified models by utilizing a fully discrete diffusion modeling to handle inputs and outputs across various modalities. This innovative approach allows Lumina-DiMOO to achieve higher sampling efficiency compared to previous autoregressive (AR) or hybrid AR-Diffusion paradigms and adeptly support a broad spectrum of multi-modal tasks, including text-to-image generation, image-to-image generation (e.g., image editing, subject-driven generation, and image inpainting, etc.), as well as image understanding. Lumina-DiMOO achieves state-of-the-art performance on multiple benchmarks, surpassing existing open-source unified multi-modal models. To foster further advancements in multi-modal and discrete diffusion model research, we release our code and checkpoints to the community. Project Page: https://synbol.github.io/Lumina-DiMOO. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_06308 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understanding Xin, Yi Qin, Qi Luo, Siqi Zhu, Kaiwen Yan, Juncheng Tai, Yan Lei, Jiayi Cao, Yuewen Wang, Keqi Wang, Yibin Bai, Jinbin Yu, Qian Jiang, Dengyang Pu, Yuandong Chen, Haoxing Zhuo, Le He, Junjun Luo, Gen Li, Tianbin Hu, Ming Ye, Jin Ye, Shenglong Zhang, Bo Xu, Chang Wang, Wenhai Li, Hongsheng Zhai, Guangtao Xue, Tianfan Fu, Bin Liu, Xiaohong Qiao, Yu Liu, Yihao Computer Vision and Pattern Recognition We introduce Lumina-DiMOO, an open-source foundational model for seamless multi-modal generation and understanding. Lumina-DiMOO sets itself apart from prior unified models by utilizing a fully discrete diffusion modeling to handle inputs and outputs across various modalities. This innovative approach allows Lumina-DiMOO to achieve higher sampling efficiency compared to previous autoregressive (AR) or hybrid AR-Diffusion paradigms and adeptly support a broad spectrum of multi-modal tasks, including text-to-image generation, image-to-image generation (e.g., image editing, subject-driven generation, and image inpainting, etc.), as well as image understanding. Lumina-DiMOO achieves state-of-the-art performance on multiple benchmarks, surpassing existing open-source unified multi-modal models. To foster further advancements in multi-modal and discrete diffusion model research, we release our code and checkpoints to the community. Project Page: https://synbol.github.io/Lumina-DiMOO. |
| title | Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understanding |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.06308 |