VMDiff: Visual Mixing Diffusion for Limitless Cross-Object Synthesis
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_ | 1866914061524926464 |
|---|---|
| author | Xiong, Zeren Yu, Yue Zhang, Zedong Chen, Shuo Yang, Jian Li, Jun |
| author_facet | Xiong, Zeren Yu, Yue Zhang, Zedong Chen, Shuo Yang, Jian Li, Jun |
| contents | Creating novel images by fusing visual cues from multiple sources is a fundamental yet underexplored problem in image-to-image generation, with broad applications in artistic creation, virtual reality and visual media. Existing methods often face two key challenges: coexistent generation, where multiple objects are simply juxtaposed without true integration, and bias generation, where one object dominates the output due to semantic imbalance. To address these issues, we propose Visual Mixing Diffusion (VMDiff), a simple yet effective diffusion-based framework that synthesizes a single, coherent object by integrating two input images at both noise and latent levels. Our approach comprises: (1) a hybrid sampling process that combines guided denoising, inversion, and spherical interpolation with adjustable parameters to achieve structure-aware fusion, mitigating coexistent generation; and (2) an efficient adaptive adjustment module, which introduces a novel similarity-based score to automatically and adaptively search for optimal parameters, countering semantic bias. Experiments on a curated benchmark of 780 concept pairs demonstrate that our method outperforms strong baselines in visual quality, semantic consistency, and human-rated creativity. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_23605 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | VMDiff: Visual Mixing Diffusion for Limitless Cross-Object Synthesis Xiong, Zeren Yu, Yue Zhang, Zedong Chen, Shuo Yang, Jian Li, Jun Computer Vision and Pattern Recognition Creating novel images by fusing visual cues from multiple sources is a fundamental yet underexplored problem in image-to-image generation, with broad applications in artistic creation, virtual reality and visual media. Existing methods often face two key challenges: coexistent generation, where multiple objects are simply juxtaposed without true integration, and bias generation, where one object dominates the output due to semantic imbalance. To address these issues, we propose Visual Mixing Diffusion (VMDiff), a simple yet effective diffusion-based framework that synthesizes a single, coherent object by integrating two input images at both noise and latent levels. Our approach comprises: (1) a hybrid sampling process that combines guided denoising, inversion, and spherical interpolation with adjustable parameters to achieve structure-aware fusion, mitigating coexistent generation; and (2) an efficient adaptive adjustment module, which introduces a novel similarity-based score to automatically and adaptively search for optimal parameters, countering semantic bias. Experiments on a curated benchmark of 780 concept pairs demonstrate that our method outperforms strong baselines in visual quality, semantic consistency, and human-rated creativity. |
| title | VMDiff: Visual Mixing Diffusion for Limitless Cross-Object Synthesis |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.23605 |