Unlocking Secrets: A Data Analysis of the Science Museum's Collection

Analyzing the Science Museum Group's open photographic dataset revealed unique items within their collection. Using computer vision techniques, researchers identified individual objects with distinctive appearances based on their distance to their five nearest neighbors. These included vintage appliances, Art Deco objects, and Kinora viewers. The analysis also highlighted trends in the colors and shapes of museum objects over time, and unexpected connections like recycled Nike shoes forming artificial turf and cullet (broken glass) used in glassmaking. This study emphasizes the value of analyzing online museum collections and the potential of computer vision in uncovering and appreciating unique artifacts.