Browsing digital video
In an operational video browser (eg CMU’s Informedia), audio track, audio type, recognised text in image, closed captions, face recognition, as well as image properties, would all be bundled together
Our work concentrates on the image properties only