Display Day - beefoo/lclabs-jfp24 GitHub Wiki

Overall Project Description (~150 words)

This project’s focus is to develop an interactive, visual tool to encourage existing and new audiences to explore the Library of Congress’s online data collections. Using machine learning models to sort and "cut" images, we will develop a collage tool using the Library’s Free to Use and Reuse collection, in addition to other collections curated by the Library. You can view these additional collections by viewing the "main-collections" sheet here. The collage tool will be an interactive interface through which users can “explore” and reinvent settings in Washington, D.C. across space and time. Users will be able to create a collage-style art piece by sorting through and placing D.C.-located images featuring different people, settings, and objects, against a set of pre-selected backdrops. The goal of this interface is to engage with and introduce a younger audience (late elementary to early middle school) to the Library through educational play.

ID-specific description (~150 words)

SH-specific description (~150 words)

While each Junior Fellow participated in every aspect of the project, we each took on an "unofficial" leadership role concerning the three main tasks of our project. These three tasks are as follows:

  1. Collection development
  2. Machine learning research and experimentation
  3. Prototyping and UI design

I worked primarily with collection development. This entailed looking through the Library's collections and working with the other two Junior Fellows to decide which collections we would like to work with and which collections would both add to the narrative of the tool and work with our chose object detection model. A large part of my work with collection development entailed working with another Junior Fellow to access the Library's API and write a script to save and download images filtered by place and rights restrictions; and save metadata associated with the collection items. I then combined wit

AP-specific description (~150 words)