Collaging with the Library’s Digital Collections

In 2018, the Library launched the very popular Free To Use and Reuse Sets, where staff curate thematic sets of items from our digital collections that are either in the public domain, have no known copyright restrictions or have been cleared by the copyright owner for public use. The public is not only free to enjoy and consume this material, but they are also encouraged to transform them, recontextualize them, and make them their own. One of the first examples of “reuse” that we nearly all encounter at a young age is collage—the art of selecting, cutting, pasting, and rearranging found material and imagery. 

In the summer of 2024, Junior Fellows Ilayda Dogan, Shauna-Kay Harrison and Aisaiah Pellecer worked with LC Labs to explore how they can create a new and engaging digital collage tool by leveraging emerging technologies and approaches in areas such as machine learning and computer vision. Starting with the Library’s Free to Use and Reuse items, the Junior Fellows used the latest approaches for computational techniques such as object detection and image segmentation to automatically identify subjects within a wide variety of images and “cut” them into segments that could be easily arranged and manipulated on an online digital canvas.

They had a particular interest in activating younger audiences and creating an experience that encourages self-directed learning and educational play, allowing users to engage with history and culture through creative expression. They also had a particular emphasis on the responsible use of machine learning technologies on Library materials, which included in-depth evaluations of the benefits, risks, and possible harms of using those technologies.

The impressive list of deliverables created by Dogan, Harrison, and Pellecer include: (1) comprehensive documentation of the creative, intellectual, and technical aspects of the project, (2) a series of Jupyter notebooks that walk users through the code that powers the project, (3) an open source web interface that allows users to create their own collage using Free to Use and Reuse collection items, and (4) evaluations of the machine learning technologies they used, including assessments of their benefits, potential risks, and risk mitigation strategies using the LC Labs AI Planning Framework.

The LC Labs team hopes this project can act as a model for how library practitioners can leverage emerging technologies and techniques in a responsible and meaningful way – as well as inspire the next generation of creators.