A Crowdsourced Open-Source Kazakh Speech Corpus and Initial Speech Recognition Baseline

Khassanov, Yerbolat; Mussakhojayeva, Saida; Mirzakhmetov, Almas; Adiyev, Alen; Nurpeiissov, Mukhamet; Varol, Hüseyin Atakan

doi:10.48550/arxiv.2009.10334

Cited by 5 publications

(12 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Kazakh speech data set uses KSC [ 34 ], which contains about 330 h of Kazakh speech data. In this paper, speech data with different time length settings are randomly selected as fine-tuning data, and the verification and test sets of the divided standards are used.…”

Section: Methodsmentioning

confidence: 99%

A Study of Speech Recognition for Kazakh Based on Unsupervised Pre-Training

Meng

Yolwas

2023

Sensors

View full text Add to dashboard Cite

Building a good speech recognition system usually requires a lot of pairing data, which poses a big challenge for low-resource languages, such as Kazakh. In recent years, unsupervised pre-training has achieved good performance in low-resource speech recognition, but it is rarely used in Kazakh and other Central and West Asian languages. In this paper, wav2vec2.0 is improved by integrating a Factorized TDNN layer to better preserve the relationship between the voice and the time step before and after the quantization, which is called wav2vec-F. The unsupervised pre-training strategy was used to learn the potential speech representation from a large number of unlabeled audio data and was applied to the cross-language ASR task, which was optimized using the noise contrast binary classification task. At the same time, speech synthesis is used to promote the performance of speech recognition. The experiment shows that wav2vec-F can effectively utilize the unlabeled data from non-target languages, and the multi-language pre-training is obviously better than the single-language pre-training. The data enhancement method using speech synthesis can bring huge benefits. Compared with the baseline model, Librispeech’s test-clean dataset has an average reduction of 1.9% in the word error rate. On the Kazakh KSC test set, the pre-training using only Kazakh reduced the word error rate by 3.8%. The pre-training of a small amount of Kazakh speech data synthesized by multi-language combined with TTS achieved a word error rate of 8.6% on the KSC test set when the labeled data were only 10 h, which was comparable to the results of the previous end-to-end model when the labeled data were 30 times less.

show abstract

Section: Methodsmentioning

confidence: 99%

A Study of Speech Recognition for Kazakh Based on Unsupervised Pre-Training

Meng

Yolwas

2023

Sensors

View full text Add to dashboard Cite

show abstract

“…However, this 300 h of the Kazakh speech corpus is not publicly available and is small for training robust E2E ASR models. To solve these limitations, Khassanov et al [ 46 ] created an open-source speech corpus for the Kazakh language that contains approximately 332 h of transcribed audio and over 153,000 utterances uttered by people of all ages and genders, and several geographical locations. They began by extracting Kazakh textual data from a variety of sources, including legislation, electronic publications, and websites, such as Wikipedia and blogs.…”

Section: Related Workmentioning

confidence: 99%

Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language

Mukhamadiyev

Khujayorov

Djuraev

et al. 2022

Sensors

View full text Add to dashboard Cite

Communication has been an important aspect of human life, civilization, and globalization for thousands of years. Biometric analysis, education, security, healthcare, and smart cities are only a few examples of speech recognition applications. Most studies have mainly concentrated on English, Spanish, Japanese, or Chinese, disregarding other low-resource languages, such as Uzbek, leaving their analysis open. In this paper, we propose an End-To-End Deep Neural Network-Hidden Markov Model speech recognition model and a hybrid Connectionist Temporal Classification (CTC)-attention network for the Uzbek language and its dialects. The proposed approach reduces training time and improves speech recognition accuracy by effectively using CTC objective function in attention model training. We evaluated the linguistic and lay-native speaker performances on the Uzbek language dataset, which was collected as a part of this study. Experimental results show that the proposed model achieved a word error rate of 14.3% using 207 h of recordings as an Uzbek language training dataset.

show abstract

“…In 2020, researchers at Nazarbayev University in Kazakhstan published a 330 h Kazakh speech corpus (KSC) [114], which was the largest Kazakh database at that time. Motivated by Common Voice, this database was also collected by crowdsourcing.…”

Section: Ksc/ksc2mentioning

confidence: 99%

“…However, they did not publish the platform, nor the data. In 2021 also, researchers from Nazarbayev University released the first large-scale Kazakh database, KSC [114], offering the first open benchmark for Kazakh speech recognition research. Since then, research on Kazakh has been in the fast lane.…”

Section: Ksc/ksc2mentioning

confidence: 99%

See 1 more Smart Citation

Automatic Speech Recognition for Uyghur, Kazakh, and Kyrgyz: An Overview

Maimaitiyiming

Nijat

et al. 2022

Applied Sciences

View full text Add to dashboard Cite

With the emergence of deep learning, the performance of automatic speech recognition (ASR) systems has remarkably improved. Especially for resource-rich languages such as English and Chinese, commercial usage has been made feasible in a wide range of applications. However, most languages are low-resource languages, presenting three main difficulties for the development of ASR systems: (1) the scarcity of the data; (2) the uncertainty in the writing and pronunciation; (3) the individuality of each language. Uyghur, Kazakh, and Kyrgyz as examples are all low-resource languages, involving clear geographical variation in their pronunciation, and each language possesses its own unique acoustic properties and phonological rules. On the other hand, they all belong to the Altaic language family of the Altaic branch, so they share many commonalities. This paper presents an overview of speech recognition techniques developed for Uyghur, Kazakh, and Kyrgyz, with the purposes of (1) highlighting the techniques that are specifically effective for each language and generally effective for all of them and (2) discovering the important factors in promoting the speech recognition research of low-resource languages, by a comparative study of the development path of these three neighboring languages.

show abstract

A Crowdsourced Open-Source Kazakh Speech Corpus and Initial Speech Recognition Baseline

Cited by 5 publications

References 11 publications

A Study of Speech Recognition for Kazakh Based on Unsupervised Pre-Training

A Study of Speech Recognition for Kazakh Based on Unsupervised Pre-Training

Automatic Speech Recognition Method Based on Deep Learning Approaches for Uzbek Language

Automatic Speech Recognition for Uyghur, Kazakh, and Kyrgyz: An Overview

Contact Info

Product

Resources

About