1 paper · 1 filter
Florian Kittler, Sheethal Bhat, Andreas Maier
Zero-shot vision-language models (VLMs) have shown promise for chest radiograph classification, but their performance is often limited by confounding label co-occurrence, long-tail…