Cross-Language End-to-End Speech Recognition Research Based on Transfer Learning for the Low-Resource Tujia Language

Yu, Chongchong; Chen, Yunbing; Li, Yueqiao; Kang, Mengzhen; Xu, Shixuan; Liu, Xueer

doi:10.3390/sym11020179

Cited by 21 publications

(15 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent developments in the automatic speech recognition field, mainly fueled by the advancements made by deep learning methods, have shown impressive results even for under-resourced languages [9,10]. Based on these results we believe that the state-of-the-art deep learning methods can be used to create a high accuracy ASR system for the Lithuanian language.…”

Section: Introductionmentioning

confidence: 77%

Lithuanian Speech Recognition Using Purely Phonetic Deep Learning

Pipiras¹,

Maskeliūnas

Damaševičius

2019

Computers

View full text Add to dashboard Cite

Automatic speech recognition (ASR) has been one of the biggest and hardest challenges in the field. A large majority of research in this area focuses on widely spoken languages such as English. The problems of automatic Lithuanian speech recognition have attracted little attention so far. Due to complicated language structure and scarcity of data, models proposed for other languages such as English cannot be directly adopted for Lithuanian. In this paper we propose an ASR system for the Lithuanian language, which is based on deep learning methods and can identify spoken words purely from their phoneme sequences. Two encoder-decoder models are used to solve the ASR task: a traditional encoder-decoder model and a model with attention mechanism. The performance of these models is evaluated in isolated speech recognition task (with an accuracy of 0.993) and long phrase recognition task (with an accuracy of 0.992).

show abstract

Section: Introductionmentioning

confidence: 77%

Lithuanian Speech Recognition Using Purely Phonetic Deep Learning

Pipiras¹,

Maskeliūnas

Damaševičius

2019

Computers

View full text Add to dashboard Cite

show abstract

“…To deal with this language we chose a program called PyCharm, a development program that supports Python language. We relied on GPU for training which has a great ability to handle matrixes and repetitive operations (Yu et al, 2019).…”

Section: Resultsmentioning

confidence: 99%

Cascade Deep Neural Networks Classifiers for Phonemes Recognition

Smit¹,

Al-Assimi²

2020

J. Eng. Appl. Sci.

View full text Add to dashboard Cite

In the last few years, deep neural networks have taken the problem of automated voice recognition to a whole new level of accuracy. Where it provided the highest recognition rates whether on words or on phonemes. Voice recognition problem represents the first phase of automated speech recognition systems. In this research, we introduce the recognition of phonemes based on deep neural networks using the Convolutional Neural Network 'CNN'. We will discuss two approaches of recognition, the direct approach by recognizing the phonemes using a single classification phase by obtaining the correct phonemes directly through the input. The second proposed approach uses several phases of classification by taking into account the types of phonemes and their classes (vowels, semi-vowels, explosive, etc.). In both approaches, we rely on the mel spectrogram transform where the acoustic signal is converted into a two-dimensional matrix within the frequency domain, this matrix is then inserted as the input of the deep neural network. We tested the proposed classifier on TIMIT database, obtained 57% accuracy in the direct approach and a higher accuracy of 61% using our proposed approach.

show abstract

“…Wang et al [44] combined CTC with Tibetan linguistics knowledge and used bound triphones as a modeling unit to solve the problem of Tibetan acoustic modeling under resource constraints, which made the recognition rate based on the end-to-end acoustic model method exceed the speech recognition system based on BLSTM-HMM. Yu et al [21], [32] used the BLSTM-CTC joint model to achieve phonemic level speech recognition for a few hours of data.…”

Section: B Acoustic Models With End-to-end Structurementioning

confidence: 99%

“…Finally, the transferred model is retrained using target low-resource language. In this case, the pre-training model can be regarded as a feature extractor [32]. In practice, the pre-training model can be used in combination with multilingual training.…”

Section: ) Transfer Learningmentioning

confidence: 99%

Acoustic Modeling Based on Deep Learning for Low-Resource Speech Recognition: An Overview

Kang

Chen

et al. 2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

The polarization of world languages is becoming more and more obvious. Many languages, mainly endangered languages, are of low-resource attribute due to lack of information. Both language conservation and cultural heritage face important challenges. Therefore, speech recognition for lowresource scenario has become a hot topic in the field of speech. Based on the complex network structures and huge model parameters, deep learning has become a powerful science in the process of speech recognition, which has a broad and far-reaching significance for the study of low-resource speech recognition. Aiming at the characteristic of low resource, this paper reviews the history and research status of two kinds of acoustic models of deep learning neural networks and acoustic end-to-end structures. We further elaborate on several key techniques for improving performance in the two aspects of data and model training. There are two projects for low-resource languages introduced in this paper. The possible future developments are finally pointed out. These works provide some reference for computer speech and language processing.

show abstract

Cross-Language End-to-End Speech Recognition Research Based on Transfer Learning for the Low-Resource Tujia Language

Cited by 21 publications

References 36 publications

Lithuanian Speech Recognition Using Purely Phonetic Deep Learning

Lithuanian Speech Recognition Using Purely Phonetic Deep Learning

Cascade Deep Neural Networks Classifiers for Phonemes Recognition

Acoustic Modeling Based on Deep Learning for Low-Resource Speech Recognition: An Overview

Contact Info

Product

Resources

About