HMPD: A Novel Dataset for Microplastics Classification with Digital Holography

Summary

Researchers developed HMPD, a novel digital holography dataset for training machine learning models to classify microplastic microfibers shed from synthetic textiles during washing. Automated, high-throughput detection methods like this are needed to monitor laundry-derived microfibers — one of the largest contributors to environmental microplastic loads — at scale in wastewater and environmental samples.