Over the weekend I attended the Association of Moving Image Archivist conference in Richmond and there was one panel that I thought would be interesting to the group.

Mark J. Williams of Dartmouth College is working on a project called ACTION (Audio-visual Cinematic Toolkit for Interaction, Organization, and Navigation) which provides a new platform to machine read film frame by frame and graph the data. The computing and math they are using are a bit lost on me, and I do not think I even understand the graphs but it seemed interesting to me that people have been working to machine read visual material to see if that can tell us something about a film. I am not sure exactly what it all means, but it seems like “there is something in there.”

Here is an example: