×

The development of software components for streaming audio content filtering through the use of hidden Markov models. (Russian. English summary) Zbl 1331.94038

Summary: The results of the development of efficient algorithms for streaming voice recognition using stochastic models based on the use of hidden Markov models are shown in this work. The article provides basic theoretical information for the hidden Markov model of the discrete system and the necessary parameters to define it are distinguished. Also there are three main tasks considered that need to be solved for the successful application of hidden Markov models in speech recognition systems. The algorithm of the method of Baum-Welch aimed at clarifying the parameters of the model and the Viterbi algorithm of selection of the most likely sequence of states of the system are given. These two methods are implemented in the environment of graphical programming LabVIEW in the form of software modules that implement the construction of the hidden Markov models of individual words, using the method of Baum-Welch and recognition of these words on the basis of the Viterbi method. It is supposed to use these modules to implement streaming audio content filtering in digital communication systems.

MSC:

94A12 Signal theory (characterization, reconstruction, filtering, etc.)
68T10 Pattern recognition, speech recognition

Software:

LabVIEW

References:

[1] [1] Гоноровский И. С., Демин М. П., Радиотехнические цепи и сигналы, Дрофа, М., 2006, 719 с. [Gonorovskij I. S., Demin M. P., Radio Circuits and Signals, Drofa, M., 2006, 719 pp. (in Russian)] · Zbl 1204.11080
[2] [2] Сергиенко А. Б., Цифровая обработка сигналов, Питер, СПб., 2007, 750 с. [Sergienko A. B., Digital signal processing, Piter, Saint Petersburg, 2007, 750 pp. (in Russian)] · Zbl 1154.68045
[3] [3] Рабинер Л. Р., ”Скрытые Марковские модели и их применение в избранных приложениях при распознавании речи”, ТИИЭР, 77 (1989), 86–120 [Rabiner L. R., ”Hidden Markov models and their application in selected applications in speech recognition”, PIEEE, 77 (1989), 86–120] · Zbl 1200.11037
[4] [4] Narada Warakagoda, A Hybrid ANN-HMM ASR system with NN based adaptive preprocessing, M. Sc. thesis,
[5] [5] Портал компании National Instruments Russia, · Zbl 0342.02023
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.