Label-Efficient Learning for Biomedical and Educational Applications

Recently, deep learning-based methods achieved promising performance in a variety of applications areas to include biomedicine and education. However, training deep learning-based methods requires a large amount of annotated data, which is time-consuming and labor-intensive. In this seminar, Dr. Don Brown of UVA Data Science will describe two approaches to this problem. One developed for educational applications uses Large Language Models (LLM) to label teacher questions and the second uses a new Label-efficient Contrastive learning-based (LECL) model to detect and classify various types of nuclei in 3D immunofluorescent images.

Watch on YouTube Live.

Learn more about the Biomedical Data Science Innovation Lab and Seminar Series.

Filed Under: Events