Kazakh Speech Research Articles

Application of the conformer model for kazakh speech recognition

Application of the conformer model for kazakh speech recognition

Comparing Machine Learning Models to Determine the Effect of Speech Duration on Speaker Identification within Kazakh Speech Corpus

This paper conducts comparative analysis of speaker identification techniques, namely Gaussian Mixture Models (GMM), Bidirectional Long Short-Term Memory (BiLSTM) neural networks, and Recurrent Neural Networks (RNN), applied to the Kazakh Speech Corpus. The study's central objective is to explore the influence of speech duration on the efficacy of speaker identification. Our results indicate that all models exhibit enhanced accuracy with increased speech duration. Furthermore, the incorporation of delta and fundamental frequency (F0) as supplementary features bolsters the performance of these models. This research contributes to the understanding of how speech duration and additional features can optimize speaker identification techniques, providing valuable insights for future developments in the field.

Cascade Speech Translation for the Kazakh Language

Speech translation systems have become indispensable in facilitating seamless communication across language barriers. This paper presents a cascade speech translation system tailored specifically for translating speech from the Kazakh language to Russian. The system aims to enable effective cross-lingual communication between Kazakh and Russian speakers, addressing the unique challenges posed by these languages. To develop the cascade speech translation system, we first created a dedicated speech translation dataset ST-kk-ru based on the ISSAI Corpus. The ST-kk-ru dataset comprises a large collection of Kazakh speech recordings along with their corresponding Russian translations. The automatic speech recognition (ASR) module of the system utilizes deep learning techniques to convert spoken Kazakh input into text. The machine translation (MT) module employs state-of-the-art neural machine translation methods, leveraging the parallel Kazakh-Russian translations available in the dataset to generate accurate translations. By conducting extensive experiments and evaluations, we have thoroughly assessed the performance of the cascade speech translation system on the ST-kk-ru dataset. The outcomes of our evaluation highlight the effectiveness of incorporating additional datasets for both the ASR and MT modules. This augmentation leads to a significant improvement in the performance of the cascade speech translation system, increasing the BLEU score by approximately 2 points when translating from Kazakh to Russian. These findings underscore the importance of leveraging supplementary data to enhance the capabilities of speech translation systems.

The Development of a Kazakh Speech Recognition Model Using a Convolutional Neural Network with Fixed Character Level Filters

This study is devoted to the transcription of human speech in the Kazakh language in dynamically changing conditions. It discusses key aspects related to the phonetic structure of the Kazakh language, technical considerations in collecting the transcribed audio corpus, and the use of deep neural networks for speech modeling. A high-quality decoded audio corpus was collected, containing 554 h of data, giving an idea of the frequencies of letters and syllables, as well as demographic parameters such as the gender, age, and region of residence of native speakers. The corpus contains a universal vocabulary and serves as a valuable resource for the development of modules related to speech. Machine learning experiments were conducted using the DeepSpeech2 model, which includes a sequence-to-sequence architecture with an encoder, decoder, and attention mechanism. To increase the reliability of the model, filters initialized with symbol-level embeddings were introduced to reduce the dependence on accurate positioning on object maps. The training process included simultaneous preparation of convolutional filters for spectrograms and symbolic objects. The proposed approach, using a combination of supervised and unsupervised learning methods, resulted in a 66.7% reduction in the weight of the model while maintaining relative accuracy. The evaluation on the test sample showed a 7.6% lower character error rate (CER) compared to existing models, demonstrating its most modern characteristics. The proposed architecture provides deployment on platforms with limited resources. Overall, this study presents a high-quality audio corpus, an improved speech recognition model, and promising results applicable to speech-related applications and languages beyond Kazakh.

Modern language trends: norm and usus features

The article analyzes the current socio-linguistic issues of the state language. Compliance with the traditional norms of the national literary language, identification of the main trends in its development, compilation of description of new norms,collection and systematization of the ways of speech delivery in a communicative practice, in this regard, in accordance with the modern linguistic anthropocentric direction, determination, revision and clarification of the concepts of Kazakh speech culture and main scientific principles are the relevant issues. Due to the transition of the Kazakh alphabet to Latin graphics, further improvement of orthological codification of the Kazakh language has turned into one of the most acute problems in the scientific and methodological aspect. The unification and normalization of the language norm / literary norm contributes to improving the level of language culture of language users, enriching the language, as well as ensuring the full functioning of the Kazakh language as the state language. The article analyzes the relationship of the concepts of norm – language norm – literary norm, usus and the conceptual distinctive features, the peculiarities characterizing the norms, such as traditional and internal dynamic ones. The examples presented in the dictionary of erroneous uses “Soztüzer”, compiled collectively by the scientists like Sh.Kurmanbayuly, Zh.Zhakypov, T.Yermekova, O.Zhubayeva, B.Suyerkul, N.Kozhabek, A.Almenbet are traditionally analyzed. Moreover the frequency of use of a linguistic fact in determining the difference between norm and usus was taken as the main criteria.

MODERN AND TRADITIONAL KAZAKH SPEECH: COMPARATIVE ANALYSIS (according to the lexico-phraseological level)

The article discusses the features of the traditional vocabulary of the Kazakh language, their changes in vocabulary and phraseology. The language of older people living in a single Kazakh environment has been analyzed in comparison with the speech characteristics of Kazakh citizens born since 2000. As a result, it was noted that in the oral language of the younger generation, the norm of written language is often observed, the word is pronounced as it is written, and even new phenomena are emerging that are alien to the norm of the Kazakh language. It was also considered in detail that the language of young people is becoming increasingly difficult, changing to short words and phrases, and the expressiveness of the language of the generation that speaks in a short and concise manner is decreasing. It is shown that such changes in the language do not correspond to the traditional grammatical system, the speech skills of language consumers with a high level of language education, and indicate that the traditional speech of the Kazakh language has changed, separated from the national and traditional character.

Development online models for automatic speech recognition systems with a low data level

Speech recognition is a rapidly growing field in machine learning. Conventional automatic speech recognition systems were built based on independent components, that is an acoustic model, a language model and a vocabulary, which were tuned and trained separately. The acoustic model is used to predict the context-dependent states of phonemes, and the language model and lexicon determine the most possible sequences of spoken phrases. The development of deep learning technologies has contributed to the improvement of other scientific areas, which includes speech recognition. Today, the most popular speech recognition systems are systems based on an end-to-end (E2E) structure, which trains the components of a traditional model simultaneously without isolating individual elements, representing the system as a single neural network. The E2E structure represents the system as one whole element, in contrast to the traditional one, which has several independent elements. The E2E system provides a direct mapping of acoustic signals in a sequence of labels without intermediate states, without the need for post-processing at the output, which makes it easy to implement. Today, the popular models are those that directly output the sequence of words based on the input sound in real-time, which are online end-to-end models. This article provides a detailed overview of popular online-based models for E2E systems such as RNN-T, Neural Transducer (NT) and Monotonic Chunkwise Attention (MoChA). It should be emphasized that online models for Kazakh speech recognition have not been developed at the moment. For low-resource languages, like the Kazakh language, the above models have not been studied. Thus, systems based on these models have been trained to recognize Kazakh speech. The results obtained showed that all three models work well for recognizing Kazakh speech without the use of external additions.

Hybrid end-to-end model for Kazakh speech recognition

Hybrid end-to-end model for Kazakh speech recognition

A study of transformer-based end-to-end speech recognition system for Kazakh language

Today, the Transformer model, which allows parallelization and also has its own internal attention, has been widely used in the field of speech recognition. The great advantage of this architecture is the fast learning speed, and the lack of sequential operation, as with recurrent neural networks. In this work, Transformer models and an end-to-end model based on connectionist temporal classification were considered to build a system for automatic recognition of Kazakh speech. It is known that Kazakh is part of a number of agglutinative languages and has limited data for implementing speech recognition systems. Some studies have shown that the Transformer model improves system performance for low-resource languages. Based on our experiments, it was revealed that the joint use of Transformer and connectionist temporal classification models contributed to improving the performance of the Kazakh speech recognition system and with an integrated language model it showed the best character error rate 3.7% on a clean dataset.

DEVELOPING STUDENTS' LANGUAGE SKILLS THROUGH INDEPENDENT WORK

The Kazakh language course conducted in higher educational institutions for russian groups has its own specific. There is a need for further development and independent improvement of the language skills acquired within the framework of the school blueprint. Achieving average-level, above-average results requires serious qualifications. Some of the optimal approaches to language acquisition have a positive impact on students. Creative adaptive learning is one of the manifestations of language literacy requisite by modern youth. It is characterized by the fact that, as a result of its consistency, purposefulness, it can have a direct impact on the quality of teaching the subject, clearly and accurately reflects the level of knowledge and skills of the student, optimizes the assessment, gives direction to determine the extant of shape of various competencies of the student and the development of his capabilities in the future. We can also motivate to perfect knowledge of the Kazakh language by systematically providing cognitive, educational, informational tasks as independent work. Such approaches are of great importance for acquisition the Kazakh language in the audience with other languages in the direction of educating students of the intelligence and mind, the culture of Kazakh thinking and speech, the formation of personal abilities.

GAN acoustic model for Kazakh speech synthesis

Recent studies on the application of generative adversarial networks (GAN) for speech synthesis have shown improvements in the naturalness of synthesized speech, compared to the conventional approaches. In this article, we present a new framework of GAN to train an acoustic model for speech synthesis. The proposed GAN consists of a generator and a pair of agent discriminators, where the generator produces acoustic parameters taking into account linguistic parameters; and the pair of agent discriminators are introduced to improve the naturalness of the synthesized speech. We feed the agents with acoustic and linguistic parameters, thereby the agents do not only examine the acoustic distribution, but also the relationship between linguistic and acoustic parameters. Training and testing were conducted on the Kazakh speech corpus. According to the results of this research, the proposed framework of GAN improves the accuracy of the acoustic model for the Kazakh text-to-speech system.

ABOUT ORTHOEPY OF THE KAZAKH LANGUAGE

The article provides a brief overview of the scientific works of Professor S. Myrzabekov, in which the scientist’s attention is focused on the problems of the orthoepy of the Kazakh language, on the need to introduce this subject into the curriculum of secondary and higher education as a compulsory subject and preliminary training of teachers. Prof. S. Myrzabekov believes that it is necessary, first, to consider in detail the patterns of pronunciation of the Kazakh speech, having also studied the history of the problem; secondly, to strive to form a scientific research base; thirdly, to carry out a phonetic substantiation of pronunciation: fourthly, to compile a dictionary of the pronunciation of Kazakh words.

INTELLIGIBILITY OF THE KAZAKH SPEECH WHEN IT'S PROTECTED WITH COMBINED MASKING SIGNALS

The article is devoted to assessing the intelligibility of the Kazakh speech when it's masked by combined signals, including «white» noise and speech-like signals. The phonetics features of the Kazakh language have been considered taking into account the law of syngarmonism and the spectrum differences of speech in the Kazakh language and speech in the Russian language. A technique for assessing the intelligibility of the Kazakh speech when it's masked by «white» noise and speech-like signals is proposed. The aim of the work is to analyze well-known methods for speech intelligibility assessing and applying these methods to assess speech intelligibility in the Kazakh language, taking into account masking by its combined signals. Due to the fact that the use of the articulation method of assessing intelligibility for the Kazakh speech requires a dependence of intelligibility on the articulation index for this particular language (the application for the Kazakh language has not been experimentally tested), the use of the formant approach to speech intelligibility assessing will be examined in more detail. The curried out experimental studies of the spectral density of speech in the Kazakh language made it possible to obtain it's approximate dependence on the frequency and take into account the phonetic features of the Kazakh speech when assessing the security of the speech information using the formant method.

MULTIMODAL SYSTEMS FOR SPEECH RECOGNITION

In this article, we have implemented a system of multimodal recognition of Kazakh speech, based on speech and lip recognition. During the feature extraction phase, several methods have been used, such as voice activity detection (VAD), mel-frequency cepstral coefficients, perceptual linear prediction, relative perceptual linear prediction, and their first-order time derivatives. The main problems of recognition of Kazakh speech, VAD algorithms and speech segmentation, lip movement recognition are considered in the article. The description of probabilistic modelling of audiovisual speech based on coupled hidden Markov models (HMMs), information fusion methods with weight coefficients for audio and video speech modalities, and parametric representation of signals is provided. Quantitative results in multimodal recognition of continuous Kazakh speech indicate high accuracy and reliability of the automatic system. This approach has been used and compared in terms of computational time and recognition speed and gives very interesting results.

Multimodal systems for speech recognition

In this article, we have implemented a system of multimodal recognition of Kazakh speech, based on speech and lip recognition. During the feature extraction phase, several methods have been used, such as voice activity detection (VAD), mel-frequency cepstral coefficients, perceptual linear prediction, relative perceptual linear prediction, and their first-order time derivatives. The main problems of recognition of Kazakh speech, VAD algorithms and speech segmentation, lip movement recognition are considered in the article. The description of probabilistic modelling of audiovisual speech based on coupled hidden Markov models (HMMs), information fusion methods with weight coefficients for audio and video speech modalities, and parametric representation of signals is provided. Quantitative results in multimodal recognition of continuous Kazakh speech indicate high accuracy and reliability of the automatic system. This approach has been used and compared in terms of computational time and recognition speed and gives very interesting results.

Acoustic modeling for Kazakh speech synthesis

We present a new framework of generative adversarial network for training of acoustic model for speech synthesis. The proposed generative adversarial network consists of a generator and a pair of agent discriminators, where the generator predicts the acoustic features from the linguistic representation. Training and testing were carried out on the Kazakh speech corpus, which consisted of 5.6 hours of speech recording. According to the experiment results the 3.46 mean opinion score was obtained which shows an acceptable quality of speech synthesis. This approach of the acoustic model development can be applied in speech synthesis systems of the other languages.

Continuous Speech Recognition of Kazakh Language

This article describes the methods of creating a system of recognizing the continuous speech of Kazakh language. Studies on recognition of Kazakh speech in comparison with other languages began relatively recently, that is after obtaining independence of the country, and belongs to low resource languages. A large amount of data is required to create a reliable system and evaluate it accurately. A database has been created for the Kazakh language, consisting of a speech signal and corresponding transcriptions. The continuous speech has been composed of 200 speakers of different genders and ages, and the pronunciation vocabulary of the selected language. Traditional models and deep neural networks have been used to train the system. As a result, a word error rate (WER) of 30.01% has been obtained.

Quality of Speech Reproduction Using Stochastic Digital Systems of Information Transfer with its Statistical Compaction

Quality of oral speech reproduction using stochastic digital systems of information transfer at its statistical compaction is analyzed (on the example of Russian and Kazakh speeches) with the ratio of the speech signal power to the digitization and restoration noise power taken as quantitative characteristic of speech description. From an analysis of the results obtained it unambiguously follows that for analytical determination of the quality of speech message restoration using stochastic digital transfer systems it is expedient not to resort to approximations, but to use for estimation the experimentally obtained distribution of the probability of the number of successively rejected counts. The preference of the approach with alternating priority is established for a Russian speech message. In consideration of the Kazakh speech, the preference of the approach with alternating priority of the message is revealed. The approach with priority alternation in a count is the best of the three approaches being considered (with fixed priority, priority alternation in the transfer cycle Е1 and priority alternation in a count) for messages of the oral Russian and Kazakh speeches. This significant preference is explained by the influence of successive losses of more than one count on the ratio of the signal to noise power.

Implementation of the Intelligent Voice System for Kazakh

Modern speech technologies are highly advanced and widely used in day-to-day applications. However, this is mostly concerned with the languages of well-developed countries such as English, German, Japan, Russian, etc. As for Kazakh, the situation is less prominent and research in this field is only starting to evolve. In this research and application-oriented project, we introduce an intelligent voice system for the fast deployment of call-centers and information desks supporting Kazakh speech. The demand on such a system is obvious if the country's large size and small population is considered. The landline and cell phones become the only means of communication for the distant villages and suburbs. The system features Kazakh speech recognition and synthesis modules as well as a web-GUI for efficient dialog management. For speech recognition we use CMU Sphinx engine and for speech synthesis- MaryTTS. The web-GUI is implemented in Java enabling operators to quickly create and manage the dialogs in user-friendly graphical environment. The call routines are handled by Asterisk PBX and JBoss Application Server. The system supports such technologies and protocols as VoIP, VoiceXML, FastAGI, Java SpeechAPI and J2EE. For the speech recognition experiments we compiled and used the first Kazakh speech corpus with the utterances from 169 native speakers. The performance of the speech recognizer is 4.1% WER on isolated word recognition and 6.9% WER on clean continuous speech recognition tasks. The speech synthesis experiments include the training of male and female voices.

Methods for applying VAD in Kazakh speech recognition systems

This article considers the algorithm "Voice activity detection" and the using VAD algorithm in the system of Kazakh speech recognition. The paper presents a mathematical model VAD and methods for d...