Christan Grant Research Group

University of Florida

How can interactive methods improve data acquisition and labeling quality for downstream models? The group develops tools and workflows for data acquisition and labeling across the data pipeline to reduce manual effort and error. Researchers build and evaluate interactive machine learning interfaces to support human-in-the-loop model training and refinement. The team applies visualization techniques to inspect and debug datasets and model outputs to support iterative dataset curation. The group investigates database and data management mechanisms to scale data-intensive experiments and reproducible analyses for data science workloads.