Yazar "Topuz, Vedat" seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim Yok
    Öğe
    A phoneme-based approach for eliminating out-of-vocabulary problem of Turkish speech recognition using Hidden Markov Model
    (C R L Publishing LTD, 2018) Yavuz, Erdem; Topuz, Vedat
    Since Turkish is a morphologically productive language, it is almost impossible for a word-based recognition system to be realized to completely model Turkish language. Due to the fact that it is difficult for the system to recognize words not introduced to it in a word-based recognition system, recognition success rate drops considerably caused by out-of-vocabulary words. In this study, a speaker-dependent, phoneme-based word recognition system has been designed and implemented for Turkish Language to overcome the problem. An algorithm for finding phoneme-boundaries has been devised in order to segment the word into its phonemes. After the segmentation of words into phonemes, each phoneme is separated into different sub-groups according to its position and neighboring phonemes in that word. Generated sub-groups are represented by Hidden Markov Model, which is a statistical technique, using Mel-frequency cepstral coefficients as feature vector. Since phoneme-based approach is adopted in this study, it has been successfully achieved that many out of vocabulary words could be recognized.
  • Küçük Resim Yok
    Öğe
    Recognition of Turkish vowels by probabilistic neural networks using Yule-Walker AR method
    (2010) Yavuz, Erdem; Topuz, Vedat
    In this work, recognition of vowels in Turkish Language by probabilistic neural networks is implemented using a spectral analysis method. Power spectral density of the phones obtained from speakers is estimated. Then weighted power spectrum is calculated after power spectral density of that phone is passed through a number of band pass filters. In this way, estimated power spectrums of the phones which are obtained from speakers approximate to a mel scale. Mel scale coefficients obtained, form the feature vector of the phone that is pronounced. These feature vectors constitute the database of the related speaker. Thus and so, every speaker has its own database. When it comes to recognize a phone pronounced by a speaker later, a probabilistic neural network model is created using the database belonging to that speaker. The feature vector of the phone which is to be recognized is computed as mentioned above. In this study, speaker-dependent recognition of Turkish vowels has been realized with an accuracy rate of over 95 percent. © 2010 Springer-Verlag.