Description of the project:
“Building a database of images and keywords for the Mead Collection (collaboration with Dr. Miloslava Hruba). This project required an interdisciplinary approach from Computer Science Department and the Mead Art museum. The initial inquiries were:
- Explore the use and functionality of Google Vision API on pre-defined Mead Art Museum collection image metadata.
- Create a program that would filter Google Vision results to match controlled language keywords and themes embedded in Mead’s existing database.
The outcome of this collaborative experiment can greatly enhance the accessibility and searchability of the museum's database and encourage science and technology-minded students and faculty to explore the art collection in an interdisciplinary fashion.”
Please refer to the attached poster for a detailed explanation of the project, including methods, results, and findings: SURF_Poster_Final.pdf
Resources
Link to the Collections Database for Five Colleges and Historic Deerfield Museum Consortium: https://museums.fivecolleges.edu/
Users can filter their search according to their needs and research. For example, this link contains all the images in Mead's database that were manually tagged “dogs.” When a web object name starts with AC, it belongs to Amherst College’s collection.