Узбек тили нуткий сузларини танишнинг яширин марков модели
Аннотация
В этой статье, с применением Марковских моделей разработана модель распознавания узбекских слов. Приведены этапы обработки речевых сигналов, решаются задачи параметризации речевых сигналов и их группировки. Подробно описан алгоритм работы Марковских цепей при распознавании отдельных слов. Проведен ряд эксперименты по распознаванию узбекской речи.
Ушбу маколада Марков моделининг кулланилиши натижасида узбек тили сузларини аниклаш модел структураси ишлаб чикилган. Нуткий сигналларни кайта ишлаш боскичлари келтирилган. Нуткий сузларни параметрларга ажратиш ва уларни гурухлаш масалалари ечилган. Сузларни аниклаш учун Марков моделнинг ишлаш принципи кенг ёритилган. Узбек тили нуткий сузларни аниклаш устида тажрибалар утказилган.
In this article, as a result of application of Markov models the recognition model Uzbek words is developed. Today, speech signals are widely used in recognation of hardware systems. Such speech recognition sytems create the possibility recognizing speech signals in several languages. Analysing of their hardware’s software and models is one of the most complex issues. Also, studying and analysing the principles of hardware and software processing of speech signals is being required. Besides, developping mathematical models of speech processing of the Uzbek language, creating of their algorithms and analyzing is one of the important tasks of the research. Principle of recognizing two and three-syllable words in Uzbek language by based on parameterization was performed. Problems of parametrization of speech signals and their group are solved. It is given stages of processing of speech signals. It is detailed described the principle of work of Markov chains when determining words. In the Markov model, the parameters of speech signals as incoming values were studied. Study for the two-syllable words found in 20 out of 17 and 3 of them are vaguely defined. Each found vaguely defined word has an error. In this study, all the three-syllable words were completely clarified. Based on the hidden Markov model, the degree of detection in the processing of the words in the Uzbek language was 80-85%. Thus, the proposed model for processing of Uzbek language can be considered as effective.
Ушбу маколада Марков моделининг кулланилиши натижасида узбек тили сузларини аниклаш модел структураси ишлаб чикилган. Нуткий сигналларни кайта ишлаш боскичлари келтирилган. Нуткий сузларни параметрларга ажратиш ва уларни гурухлаш масалалари ечилган. Сузларни аниклаш учун Марков моделнинг ишлаш принципи кенг ёритилган. Узбек тили нуткий сузларни аниклаш устида тажрибалар утказилган.
In this article, as a result of application of Markov models the recognition model Uzbek words is developed. Today, speech signals are widely used in recognation of hardware systems. Such speech recognition sytems create the possibility recognizing speech signals in several languages. Analysing of their hardware’s software and models is one of the most complex issues. Also, studying and analysing the principles of hardware and software processing of speech signals is being required. Besides, developping mathematical models of speech processing of the Uzbek language, creating of their algorithms and analyzing is one of the important tasks of the research. Principle of recognizing two and three-syllable words in Uzbek language by based on parameterization was performed. Problems of parametrization of speech signals and their group are solved. It is given stages of processing of speech signals. It is detailed described the principle of work of Markov chains when determining words. In the Markov model, the parameters of speech signals as incoming values were studied. Study for the two-syllable words found in 20 out of 17 and 3 of them are vaguely defined. Each found vaguely defined word has an error. In this study, all the three-syllable words were completely clarified. Based on the hidden Markov model, the degree of detection in the processing of the words in the Uzbek language was 80-85%. Thus, the proposed model for processing of Uzbek language can be considered as effective.