In this work we present a novel two-layer hybrid Graphical model for combined shot and scene boundary detection in videos. In the first layer of the model, low-level features are...
Marc Al-Hames, Stefan Zettl, Frank Wallhoff, Steph...
There are a number of shots in a video, each of which has boundary types, such as cut, fade, dissolve and wipe. Many previous approaches can find the cut boundary without difficul...
In extended video sequences, individual frames are grouped into shots which are defined as a sequence taken by a single camera, and related shots are grouped into scenes which are...
Digital video applications exploit the intrinsic structure of video sequences. In order to obtain and represent this structure for video annotation and indexing tasks, the main ini...
In this paper, we describe a statistical method to detect highlights in a baseball game video. The input video is first segmented into scene shots, within which the camera motion ...