Speech vs nonspeech segmentation of audio signals using support vector machines

Danisman T., ALPKOÇAK A.

IEEE 15th Signal Processing and Communications Applications Conference, Eskişehir, Turkey, 11 - 13 June 2007, pp.854-857 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/siu.2007.4298688
  • City: Eskişehir
  • Country: Turkey
  • Page Numbers: pp.854-857
  • Keywords: speech segmentation, statistical signal processing
  • Dokuz Eylül University Affiliated: Yes

Abstract

In this study, we have presented a speech vs nonspeech segmentation of audio signals extracted from video. We have used 4330 seconds of audio signal extracted from "Lost" TV series for training. Our training set is automatically builded by using timestamp information exists in subtitles. After that, silence areas within those speech areas are discarded with a further study. Then, standard deviation of MFCC feature vectors of size 20 have been obtained Finally, Support Vector Machines (SVM) is used with one-vs-all method for the classification.