Christopher R. Cooney

PlumageParts provides a dataset of 4,705 bird images annotated for nine plumage regions and benchmarks segmentation models, achieving 84.01% mean Intersection over Union with a frozen DINOv3 encoder.

The paper uses citizen science and deep learning to show that losing threatened bird species will disproportionately reduce avian plumage diversity, especially in tropical regions.

The paper investigates how historical contingency affects feather nanostructure diversity in cuckoos, finding that specialized melanosomes are present in both metallic and drab plumage, suggesting retained traits can facilitate future evolution of metallic luster.

The paper examines whether sexual selection predicts population trends in birds using abundance data across 738 species and finds no significant association.