Automatic Score Scene Detection for Baseball Video

Koichi Shinoda, Kazuki Ishihara, Sadaoki Furui, Takahiro Mochizuki

Abstract: We propose a robust score scene detection method for baseball broadcast videos. This method is based on the data-driven approach which has been successful in statistical speech recognition. Audio and video feature streams are integrated by a multi-stream hidden Markov model to model each scene. The proposed method was evaluated in score scene detection experiments using video data of 25 baseball games. While the recall rate with video mode only was 82.8% and that with audio mode only was 86.6%, the proposed method achieved 90.4%. This method was proved to be significantly effective to reduce the cost for making highlight for baseball video content.