论文信息 - Topic segmentation and retrieval system for lecture videos based on spontaneous speech recognition

Topic segmentation and retrieval system for lecture videos based on spontaneous speech recognition

In this paper, we propose a segmentation method of continuous lecture speech into topics. A lecture includes several topics but it is difficult to judge their boundaries. To solve this problem, transcriptions obtained by spontaneous speech recognition of a lecture speech is associated with the textbook used in the lecture. This method showed high performance of the topic segmentation with an average of 93.7%. Incorporating this method, we constructed a system where we can view an interesting part of lecture videos, by specifying the chapters or sections as well as keywords.

Yasuo Ariki | Jun Ogata | Natsuo Yamamoto

[1] Gerald Salton,et al. Automatic text processing , 1988 .

[2] Hitoshi Isahara,et al. Toward the realization of spontaneous speech recognition - introduction of a Japanese priority program and preliminary results - , 2000, INTERSPEECH.

[3] Hermann Ney,et al. A word graph algorithm for large vocabulary continuous speech recognition , 1994, Comput. Speech Lang..

[4] Shuichi Itahashi,et al. The design of the newspaper-based Japanese large vocabulary continuous speech recognition corpus , 1998, ICSLP.

[5] Philip C. Woodland,et al. Maximum likelihood linear regression for speaker adaptation of continuous density hidden Markov models , 1995, Comput. Speech Lang..

[6] Yasuo Ariki,et al. An efficient lexical tree search for large vocabulary continuous speech recognition , 2000, INTERSPEECH.