The Medieval Department

Year
2024
Category
Machine Learning
Role
Designer and Developer

The Medieval Art department at The Metropolitan Museum of Art comprises over 4,000 individual objects. This personal project involved creating an interactive 3D embedding plot of The Met's Medieval Art collection using ResNet50 and Three.js.

ResNet50 was employed to generate feature embeddings for each piece, producing unique visual "fingerprints" that reveal patterns across the collection. These embeddings were then mapped into a 3D space with Three.js, enabling users to explore relationships and clusters among artworks, uncovering insights into style, material, and form within the collection.

Through this visualization, users can explore patterns and connections within the collection, gaining a unique perspective on the stylistic and material relationships among medieval artworks.