Cinematic Common Sense

With the proliferation of cheap video cameras that are always on, and always ready to record, distinctions between the amateur and professional videographer become blurred. When we can record at any moment, how will we decide what we want to record, and when and where we need to be in order to capture what we want? How can we best index the recorded material to support later communication needs? In this research, we use a common-sense approach to knowledge and inference to support a partnership between camera and human videographer. We present a new paradigm for producing common-sense video metadata and we show how the metadata can have a positive impact on video content capture, representation, and presentation.