Model-based Cleaning of the QUILT-1M Pathology Dataset for Text-Conditional Image Synthesis

arXiv:2404.07676v1 Announce Type: new
Abstract: The QUILT-1M dataset is the first openly available dataset containing images harvested from various online sources. While it provides a huge data variety, the image quality and composition is highly heterogeneous, impacting its utility for text-conditional image synthesis. We propose an automatic pipeline that provides predictions of the most common impurities within the images, e.g., visibility of narrators, desktop environment and pathology software, or text within the image. Additionally, we propose to use semantic alignment …

abstract arxiv cleaning cs.ai cs.cv data dataset image images openly pathology pipeline quality synthesis text type utility