I have been working with colleagues from around MSU on distant visualization of imagery for several years, and across multiple mediums. The technical-oriented part of our practice focuses on looking across large sets of images from a single film or collection, making observations based off of the analysis via the computer more quickly and deeply than with traditional methods. We use a variety of open source tools including ImageJ and ShotDetect for analysis, and have developed several scripts and methods for preparing data for analysis. These images and analysis methods are then shared via our immersive 360 room, as well other spaces for public display.
This work has focused on practical applications of large-scale image analysis. We do this work in conversation with text analysis and annotation as digital humanities methods writ large. A core area of work is teaching students these methods as a way of developing future scholars and to support faculty work in utilizing these tools in their research. Some examples of this student work in practice includes analyzing a large set of images from the Detroit Institute of Arts to recognize movements in painting, analyzing how movie cuts are organized to influence the way the movie is viewed, analyzing gameplay of video games posted on YouTube, and an analysis of the Pure Michigan campaign Instagram account post content over a number of years.
Mapes, K. & Schopieray, S. (2018). Image Analysis at Scale in the Undergraduate Humanities Classroom. Chicago Digital Humanities and Computer Science Colloquium, Chicago, IL.
Bering-Porter, D., Schopieray, S., Walker, T., Mejeur, C., Charley, M., & Bills, P. (2015). Distant Reading Visual Media Using Computation and Digital Image Analysis Tools. HASTAC Annual Conference, East Lansing, MI.
Posts on our Work