Abstract
The rise of generative models has led to increased use of large-scale datasets collected from the internet, often with minimal or no data curation. This raises concerns about the inclusion of sensitive or private information. In this work, we explore the presence of pregnancy ultrasound images, which contain sensitive personal information and are often shared online. Through a systematic examination of LAION-400M dataset using CLIP embedding similarity, we retrieve images containing pregnancy ultrasound and detect thousands of entities of private information such as names and locations. Our findings reveal that multiple images have high-risk information that could enable re-identification or impersonation. We conclude with recommended practices for dataset curation, data privacy, and ethical use of public image datasets.
Abstract (translated)
生成模型的兴起导致了大规模数据集的广泛使用，这些数据集通常是从互联网上收集而来，并且很少或根本没有进行数据整理。这引发了对敏感或私人信息被包含在内的担忧。在这项工作中，我们探讨了妊娠超声图像的存在情况，这类图像是在网上经常分享的并包含了高度个人化的敏感信息。通过系统地检查LAION-400M数据集中的CLIP嵌入相似性，我们检索到了含有妊娠超声图像的内容，并检测到数千个私人信息实例，如姓名和地点。我们的研究发现表明，多个图像中存在高风险的信息，这些信息可能使重新识别或冒充成为可能。最后，我们提出了关于数据整理、数据隐私以及公共图像数据集的伦理使用方面的建议实践方法。
URL
https://arxiv.org/abs/2602.07149
