Template-based Keyword Search with pseudo posteriorgrams Yapay Posteriorgramlar ile Şablon Temelli Anahtar Sözcük Arama


Gündoǧdu B., Sari L., Çetinkaya G., SARAÇLAR M.

24th Signal Processing and Communication Application Conference, SIU 2016, Zonguldak, Türkiye, 16 - 19 Mayıs 2016, ss.973-976, (Tam Metin Bildiri)

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/siu.2016.7495904
  • Basıldığı Şehir: Zonguldak
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.973-976
  • Anahtar Kelimeler: keyword search, low-resource languages, subsequence dynamic time warping
  • Boğaziçi Üniversitesi Adresli: Evet

Özet

In this work, a template-based search approach is adopted for the Keyword Search (KWS) problem on two of the low-resource languages (Turkish and Swahili). In low-resource languages, the use of Large Vocabulary Continuous Speech Recognition (LVCSR) systems in KWS tasks may perform poorly especially on out-of-vocabulary words. In the proposed method, the keywords are modeled to be in the same form of the audio search document in an artificial manner and utilizing template-based search methods, the performance of the baseline system is improved.