• search hit 49 of 634
Back to Result List

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

  • 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 filtering of the image-text pairs. Our findings demonstrate that by rigorously filtering the dataset, there is a substantial enhancement of image fidelity in text-to-image tasks.

Download full text files

Export metadata

Additional Services

Search Google Scholar
Metadaten
Author:Marc AubrevilleORCiD, Jonathan GanzORCiD, Jonas AmmelingORCiD, Christopher C. Kaltenecker, Christof BertramORCiD
Language:English
Document Type:Preprint
Year of first Publication:2024
Publisher:arXiv
Place of publication:Ithaca
Pages:4
Review:nein
Open Access:ja
URN:urn:nbn:de:bvb:573-48648
Related Identifier:https://doi.org/10.48550/arXiv.2404.07676
Faculties / Institutes / Organizations:Fakultät Informatik
AImotion Bavaria
Licence (German):License Logo Creative Commons BY 4.0
Release Date:2024/07/09