Automated Black-box Prompt Engineering for Personalized Text-to-Image Generation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909738162192384 |
|---|---|
| author | He, Yutong Robey, Alexander Murata, Naoki Jiang, Yiding Williams, Joshua Nathaniel Pappas, George J. Hassani, Hamed Mitsufuji, Yuki Salakhutdinov, Ruslan Kolter, J. Zico |
| author_facet | He, Yutong Robey, Alexander Murata, Naoki Jiang, Yiding Williams, Joshua Nathaniel Pappas, George J. Hassani, Hamed Mitsufuji, Yuki Salakhutdinov, Ruslan Kolter, J. Zico |
| contents | Prompt engineering is an effective but labor-intensive way to control text-to-image (T2I) generative models. Its time-intensive nature and complexity have spurred the development of algorithms for automated prompt generation. However, these methods often struggle with transferability across T2I models, require white-box access to the underlying model, or produce non-intuitive prompts. In this work, we introduce PRISM, an algorithm that automatically produces human-interpretable and transferable prompts that can effectively generate desired concepts given only black-box access to T2I models. Inspired by large language model (LLM) jailbreaking, PRISM leverages the in-context learning ability of LLMs to iteratively refine the candidate prompt distribution built upon the reference images. Our experiments demonstrate the versatility and effectiveness of PRISM in generating accurate prompts for objects, styles, and images across multiple T2I models, including Stable Diffusion, DALL-E, and Midjourney. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_19103 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Automated Black-box Prompt Engineering for Personalized Text-to-Image Generation He, Yutong Robey, Alexander Murata, Naoki Jiang, Yiding Williams, Joshua Nathaniel Pappas, George J. Hassani, Hamed Mitsufuji, Yuki Salakhutdinov, Ruslan Kolter, J. Zico Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Machine Learning Prompt engineering is an effective but labor-intensive way to control text-to-image (T2I) generative models. Its time-intensive nature and complexity have spurred the development of algorithms for automated prompt generation. However, these methods often struggle with transferability across T2I models, require white-box access to the underlying model, or produce non-intuitive prompts. In this work, we introduce PRISM, an algorithm that automatically produces human-interpretable and transferable prompts that can effectively generate desired concepts given only black-box access to T2I models. Inspired by large language model (LLM) jailbreaking, PRISM leverages the in-context learning ability of LLMs to iteratively refine the candidate prompt distribution built upon the reference images. Our experiments demonstrate the versatility and effectiveness of PRISM in generating accurate prompts for objects, styles, and images across multiple T2I models, including Stable Diffusion, DALL-E, and Midjourney. |
| title | Automated Black-box Prompt Engineering for Personalized Text-to-Image Generation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2403.19103 |