Geri Dön

A Speech synthesis system for Turkish language based on the concetanation of phonemes taken from speaker

Bir konuşmacıdan alınan fonemlerin birleştirilmesine dayanan bir Türkçe konuşma sentezleme sistemi

  1. Tez No: 29591
  2. Yazar: İLHAN YAŞAR ÖZÜM
  3. Danışmanlar: YRD. DOÇ. DR. METE BULUT
  4. Tez Türü: Yüksek Lisans
  5. Konular: Dilbilim, Elektrik ve Elektronik Mühendisliği, Linguistics, Electrical and Electronics Engineering
  6. Anahtar Kelimeler: Speech Synthesis, Waveform Concatenation, Phoneme, Diphone, Syllabification, Phonetic Transcription, Turkish Language, Speech Synthesis, Waveform Concatenation, Phoneme, Diphone, Syllabification, Phonetic Transcription, Turkish Language
  7. Yıl: 1993
  8. Dil: İngilizce
  9. Üniversite: Orta Doğu Teknik Üniversitesi
  10. Enstitü: Fen Bilimleri Enstitüsü
  11. Ana Bilim Dalı: Elektrik ve Elektronik Mühendisliği Ana Bilim Dalı
  12. Bilim Dalı: Belirtilmemiş.
  13. Sayfa Sayısı: 91

Özet

ABSTRACT A SPEECH SYNTHESIS SYSTEM FOR TURKISH LANGUAGE BASED ON THE CONCATENATION OF PHONEMES TAKEN FROM A SPEAKER ÖZÜM, İlhan Yaşar M. S. in Electrical and Electronics Engineering Supervisor: Asst. Prof. Dr. M. Mete BULUT December, 1993, 79 pages In this work, a speech synthesis system for the Turkish language has been realized. The system reads a text stored in the form of an ASCII file. The method used is based on the waveform concatenation of speech samples taken from a human speaker. The speaker utters some words and the resulting speech is digitized. The waveforms corresponding to all of the phonemes in the Turkish language are separated from this speech samples and written on disk. The separation process utilizes the first differences of the frame energy sequence. The text is then syllabified and its phonetic transcription is generated. All the syllables in the text are then synthesized by concatenating the waveforms of the phonemes forming them. As the phoneme is used as the basic unit of synthesis, the storage requirements of the system are quite low. However, the intelligibility of the resulting speech is also low because of imperfect transitions between the phoneme inwaveforms. The system also uses diphones as the units of concatenation. This decreases the noise introduced by imperfect transitions, increasing the quality of the resulting speech. This approach, however, also triples the amount of the storage used. The quality may be improved further by using some interpolation procedure at the joining boundaries of the waveforms.

Özet (Çeviri)

ABSTRACT A SPEECH SYNTHESIS SYSTEM FOR TURKISH LANGUAGE BASED ON THE CONCATENATION OF PHONEMES TAKEN FROM A SPEAKER ÖZÜM, İlhan Yaşar M. S. in Electrical and Electronics Engineering Supervisor: Asst. Prof. Dr. M. Mete BULUT December, 1993, 79 pages In this work, a speech synthesis system for the Turkish language has been realized. The system reads a text stored in the form of an ASCII file. The method used is based on the waveform concatenation of speech samples taken from a human speaker. The speaker utters some words and the resulting speech is digitized. The waveforms corresponding to all of the phonemes in the Turkish language are separated from this speech samples and written on disk. The separation process utilizes the first differences of the frame energy sequence. The text is then syllabified and its phonetic transcription is generated. All the syllables in the text are then synthesized by concatenating the waveforms of the phonemes forming them. As the phoneme is used as the basic unit of synthesis, the storage requirements of the system are quite low. However, the intelligibility of the resulting speech is also low because of imperfect transitions between the phoneme inwaveforms. The system also uses diphones as the units of concatenation. This decreases the noise introduced by imperfect transitions, increasing the quality of the resulting speech. This approach, however, also triples the amount of the storage used. The quality may be improved further by using some interpolation procedure at the joining boundaries of the waveforms.

Benzer Tezler

  1. Tez No

    33899

    PC based speech synthesis for Turkish

    Başlık çevirisi yok

    KAMİL GÜVEN

    Yüksek Lisans

    İngilizce

    İngilizce

    1994

    Elektrik ve Elektronik MühendisliğiÇukurova Üniversitesi

    Elektrik-Elektronik Ana Bilim Dalı

    YRD. DOÇ. DR. ALTUN ZİYA GÖKALP

  2. Tez No

    46192

    Ses analizi ve ses sıkıştırma ve kodlama algoritmaları

    Başlık çevirisi yok

    SERVET YILDIRIM

    Yüksek Lisans

    Türkçe

    Türkçe

    1995

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    Y.DOÇ.DR. M. ERTUĞRUL ÇELEBİ

  3. Tez No

    35223

    Türkçe ses sentezleyici

    Başlık çevirisi yok

    SELAMİ SADIÇ

    Yüksek Lisans

    Türkçe

    Türkçe

    1994

    DilbilimEskişehir Osmangazi Üniversitesi

    Elektrik-Elektronik Mühendisliği Ana Bilim Dalı

    PROF. DR. ATALAY BARKANA

  4. Tez No

    100661

    Düşük bir hızlarında konuşma kodlama ve uygulamaları

    Low bit rate speech coding and applications

    TARIK AŞKIN

    Doktora

    Türkçe

    Türkçe

    1999

    Elektrik ve Elektronik Mühendisliğiİstanbul Teknik Üniversitesi

    PROF.DR. GÜNSEL DURUSOY

  5. Tez No

    570372

    Turkish text to speech using children's voices syllables

    Çocuk ses heceleri kullanarak türkçe metinden konuşma seslendirme

    YOLDAŞ ERDOĞAN

    Yüksek Lisans

    İngilizce

    İngilizce

    2019

    Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve KontrolÇukurova Üniversitesi

    Elektrik-Elektronik Mühendisliği Ana Bilim Dalı

    DOÇ. DR. ZEKERİYA TÜFEKCİ

Geri Dön