Show abstract - PDF - Reviews

Recent advancements in zero-shot learning have enabled the use of paired image-text data to replace structured labels, replacing the need for expert annotated datasets. Domain pre-trained models, such as CXR-BERT, BlueBERT, and ClinicalBERT, offer the potential to improve the performance of CLIP-like models with specific domain knowledge by replacing BERT weights at the cost of breaking the original model\'s alignment. We evaluate the performance of zero-shot classification models with domain-specific pre-training for detecting low-prevalence pathologies. Even though replacing the weights of the original CLIP-BERT degrades model performance on commonly found pathologies, we show that pre-trained text towers perform exceptionally better on low-prevalence diseases. This motivates future ensemble models with a combination of differently trained language models for maximal performance.
Hide abstract