Researchers in Chile built a large, high-quality image library of 16 pollen types from the Biobío Region, complete with precise shape outlines verified by a pollen expert. This dataset is designed to train AI systems that can automatically identify pollen — a task that matters for tracking plant diversity, allergies, and honey authenticity.
1
The dataset contains 16,198 high-resolution microscopy images (3088×2064 pixels) with 36,383 hand-verified pollen outlines across 16 species, making it one of the most detailed pollen datasets available.
2
Each pollen grain was photographed at three focal depths, capturing surface and interior features invisible in single-plane images — a design that improved AI classification robustness.
3
A baseline AI model trained on the dataset achieved 0.985 mask mAP@50 (near-perfect detection accuracy) on the validation set after just 50 training cycles.
mail
Weekly plant science — one email, Saturdays.
Check your inbox to confirm — link expires in 24 hours.
Something went wrong — please try again.
Too many signup attempts from your network. Try again in an hour, or email hello@plant.news.
