Auto-classifying Salient Content In Video

In this position paper we present a procedure with which large volumes of video data can be automatically processed to extract only that which is salient given the current context of analysis. This technique is presented as an extension to Replayer: a software toolkit for use in the development cycle of mobile applications. By augmenting the cameras used to capture a mobile experiment we are able to establish both a location and heading for each camera, and thus generate a field of vision. Locations of each trial participant are also recorded and compared against camera visual fields to determine which periods of user activity have been recorded. A report of initial testing is given, whereby the technique’s use is demonstrated on a trivial application.