Development of an Information Retrieval System of Video-Based Learning Materials and a Method to Attach Timecode Information to Speech Data  

Oleh:

Morimoto, Yousuke ; Murota, Masao ; Shimizu, Yasutaka

Jenis: Article from Journal
Dalam koleksi: Systems and Computers in Japan vol. 38 (2007), page 1515-1524.
Topik: information retrieval; system development; learning media; speech recognition.
Fulltext: vidio based learning.pdf (994.23KB)

Isi artikel

A system that can retrieve video-based learning materials using Japanese language (SEMP: Search Engine for Motion Pictures) was developed in this research. SEMP allows full text retrieval of motion pictures using transcribed text within motion pictures and can also play back only portions that include the retrieved words. In order to provide accurate retrieval results, the required transcribed text of a motion picture is generated into text and prepared using a manual procedure by SEMP. Further, in order to play back portions that include the retrieved words, a program was developed to automatically execute an operation that attaches timecode information to the transcribed text. This program to attach timecode information uses speech recognition technology to attach timecode information to the transcribed text. We confirmed that the results of evaluation tests of the program to attach timecode information and the practical application were at high levels. © 2007 Wiley Periodicals, Inc. Syst Comp Jpn, 38(7): 53–64, 2007; Published online in Wiley InterScience (www.interscience. wiley.com). DOI 10.1002/scj.20639

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali