I wanted to create a web interface that explores a dataset using machine learning techniques. My goal was to build something that could serve as a portal into a museum collection and offer users a tool to engage with art in new ways.
This project is an exploration of how machines perceive and interpret images. Using the YOLO (You Only Look Once) object detection framework, I built a website and filtering engine for a subset of the MET's digital collection of European paintings.
YOLO is a Machine Learning framework that uses Convolutional Neural Networks (CNN) and Deep Learning for real-time object detections in images. It is trained on the COCO dataset, a large-scale object detection, segmentation, and captioning dataset that is sorted into 80 different categories of objects.
To build my interface, I used the MET's Art Collection API and downloaded a large set of images. I ran this dataset through the YOLO framework to identify the objects and their likelihood of correspondence within the artwork. To demonstrate the filtering engine at work, I built a responsive website to allow users to engage with the artworks.
