Prompt Stealing Attacks Against Text-to-Image Generation Models

Abstract

Demonstrates attacks that steal the prompts used to generate images from text-to-image models, raising IP and privacy concerns.

Categories

Framework mappings

@inproceedings{shen2024prompt,
  title = {{Prompt Stealing Attacks Against Text-to-Image Generation Models}},
  author = {Xinyue Shen and Yiting Qu and Michael Backes and Yang Zhang},
  year = {2024},
  booktitle = {USENIX Security 2024},
  eprint = {2302.09923},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2302.09923}
}