Object-level Visual Prompts for Compositional Image Generation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912175564521472 |
|---|---|
| author | Parmar, Gaurav Patashnik, Or Wang, Kuan-Chieh Ostashev, Daniil Narasimhan, Srinivasa Zhu, Jun-Yan Cohen-Or, Daniel Aberman, Kfir |
| author_facet | Parmar, Gaurav Patashnik, Or Wang, Kuan-Chieh Ostashev, Daniil Narasimhan, Srinivasa Zhu, Jun-Yan Cohen-Or, Daniel Aberman, Kfir |
| contents | We introduce a method for composing object-level visual prompts within a text-to-image diffusion model. Our approach addresses the task of generating semantically coherent compositions across diverse scenes and styles, similar to the versatility and expressiveness offered by text prompts. A key challenge in this task is to preserve the identity of the objects depicted in the input visual prompts, while also generating diverse compositions across different images. To address this challenge, we introduce a new KV-mixed cross-attention mechanism, in which keys and values are learned from distinct visual representations. The keys are derived from an encoder with a small bottleneck for layout control, whereas the values come from a larger bottleneck encoder that captures fine-grained appearance details. By mixing keys and values from these complementary sources, our model preserves the identity of the visual prompts while supporting flexible variations in object arrangement, pose, and composition. During inference, we further propose object-level compositional guidance to improve the method's identity preservation and layout correctness. Results show that our technique produces diverse scene compositions that preserve the unique characteristics of each visual prompt, expanding the creative potential of text-to-image generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_01424 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Object-level Visual Prompts for Compositional Image Generation Parmar, Gaurav Patashnik, Or Wang, Kuan-Chieh Ostashev, Daniil Narasimhan, Srinivasa Zhu, Jun-Yan Cohen-Or, Daniel Aberman, Kfir Computer Vision and Pattern Recognition Artificial Intelligence Graphics We introduce a method for composing object-level visual prompts within a text-to-image diffusion model. Our approach addresses the task of generating semantically coherent compositions across diverse scenes and styles, similar to the versatility and expressiveness offered by text prompts. A key challenge in this task is to preserve the identity of the objects depicted in the input visual prompts, while also generating diverse compositions across different images. To address this challenge, we introduce a new KV-mixed cross-attention mechanism, in which keys and values are learned from distinct visual representations. The keys are derived from an encoder with a small bottleneck for layout control, whereas the values come from a larger bottleneck encoder that captures fine-grained appearance details. By mixing keys and values from these complementary sources, our model preserves the identity of the visual prompts while supporting flexible variations in object arrangement, pose, and composition. During inference, we further propose object-level compositional guidance to improve the method's identity preservation and layout correctness. Results show that our technique produces diverse scene compositions that preserve the unique characteristics of each visual prompt, expanding the creative potential of text-to-image generation. |
| title | Object-level Visual Prompts for Compositional Image Generation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Graphics |
| url | https://arxiv.org/abs/2501.01424 |