Isolated malay speech recognition using Hidden Markov Models

Rosdi, Fadhilah; Ainon, Raja Noor

doi:10.1109/iccce.2008.4580699

Cited by 16 publications

(7 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, the ability of the neural network in recognizing isolated spoken Malay utterances in a speaker-independent manner is investigated in this paper. A lot of researches have been carried out in adult speech recognition of Malay language [10,[16][17][18]. This study uses the Malay language, which is a branch of the Austronesian (Malayo-Polynesian) language family.…”

Section: Introductionmentioning

confidence: 99%

Measuring the performance of isolated spoken Malay speech recognition using Multi-layer Neural Networks

Seman

Bakar

2010

2010 International Conference on Science and Social Research (CSSR 2010)

View full text Add to dashboard Cite

Abstract-This paper describes speech signal modeling techniques which are suited to high performance and robust isolated word recognition. In this study, a speech recognition system is presented, specifically an isolated spoken Malay word recognizer which uses spontaneous and formally speeches collected from Parliament of Malaysia. Currently the vocabulary is limited to 25 words that can be pronounced exactly as it written and controls the distribution of the vocalic segments. The speech segmentation task is achieved by adopted energy based parameter and zero crossing rate measure with modification to better locates the beginning and ending points of speech from the spoken words. The training and recognition processes are realized by using Multi-layer Perceptron (MLP) Neural Networks with two-layer network configurations that are trained with stochastic error back-propagation to adjust its weights and biases after presentation of every training data. The Melfrequency Cepstral Coefficients (MFCCs) has been chosen as speech extraction approach from each segmented utterance as characteristic features for the word recognizer. Recognition results showed that the performance of the two-layer networks increased as the numbers of hidden neurons increased. The best network structures average classification rate is 84.731% with (150-25) configuration. Implementation results also showed that the conjugate gradient (CG) algorithm was more accurate and reliable than the Levenberg-Marquardt (LM) algorithm for the network complexities and data size considered in this study.

show abstract

Section: Introductionmentioning

confidence: 99%

Measuring the performance of isolated spoken Malay speech recognition using Multi-layer Neural Networks

Seman

Bakar

2010

2010 International Conference on Science and Social Research (CSSR 2010)

View full text Add to dashboard Cite

show abstract

“…By using HMM model, they finally got an average of 88.67% recognition rate. But one drawback of this research work is that a small dataset is used here for training purpose, and for this reason, the recognition accuracy of what they achieved is too high, otherwise the determination of perfect accuracy is not so easy [5]. Four different models are introduced in automatic speech recognition system.…”

Section: Literature Reviewmentioning

confidence: 99%

Bangla Spoken Numerals Recognition by Using HMM

Paul

Adhikary

Dey

et al. 2021

Computational Intelligence in Pattern Recognition

View full text Add to dashboard Cite

Speech is one of the most natural forms of vocalized communication media. Nowadays with the advancement of machine learning, different doors are opened to us for finding several standard ways to step out in the real world. ASR is just like the door to explore the concept of communication through speech between human and digital devices that can recognize speech. In this paper, we have designed a Hidden Markov Model-based isolated Bangla numerals recognition system where the Short-Term Fourier Transform is used for collecting the feature vectors. The defined system achieved 91.50% accuracy for our own dataset of 2000 uttered samples for 10 classes, which gives a satisfied result for this Bangla numerals recognition. IntroductionStandard communication is always done successfully through speech. Speech helps us to complete a smooth communication among us. Those who are not computer professionals can communicate through speech because of its easiness and coziness in communication purposes. As the people of India live in a semi-illiterate country, so with the help of the application which supports speech recognition they will be more benifitable and can take the advantage of modern science. Speech recognition is the way that helps people to communicate with computer through speech. In speech recognition, isolated words are recognized and after converting it to text format, finally, it is prepared to a machine-readable format. Speech recognition performs a major role in medical unit, home automation, for growing the development of realworld market-based applications, which is used basically for commercial purposes. The application of speech recognition will become more useful in that field where the keyboard operation is not suitable. Speech is really needful to the users who mainly use hands and eyes for their work, like mail-sorter, aircraft pilot, cartographer, etc. as

show abstract

“…It is a type of speech feature involving coefficients that represent audio which are derived from a type of cepstral representation of the audio clip [17]. MFCC has been successfully used in recent speech processing related work such as in non-speech detection of dysarthric speech [18], isolated spoken speech recognition [13] and continuous speech recognition [19].…”

Section: A Mel Frequency Cepstral Coefficientsmentioning

confidence: 99%

Impact of acoustical voice activity detection on spontaneous filled pause classification

Hamzah

Jamil

Seman

et al. 2014

2014 IEEE Conference on Open Systems (ICOS)

View full text Add to dashboard Cite

Filled pause detection is imperative for spontaneous speech recognition as it may degrade speech recognition rate. However, filled pause is commonly confused with elongation as they shared the same acoustical properties. Few attempts of classifying filled pause and elongation employed Hidden Markov model. Our proposed method of utilizing Neural Network as a classifier achieved 96% precision rate. We also proved that voice activity detection (VAD) affects the performance of speech recognition. Three acoustical-based VAD are compared and the best precision rate is achieved by incorporating volume and firstorder difference features. Experiments are conducted using Malay language spontaneous speeches of Malaysia Parliamentary Debate sessions.

show abstract

Isolated malay speech recognition using Hidden Markov Models

Cited by 16 publications

References 4 publications

Measuring the performance of isolated spoken Malay speech recognition using Multi-layer Neural Networks

Measuring the performance of isolated spoken Malay speech recognition using Multi-layer Neural Networks

Bangla Spoken Numerals Recognition by Using HMM

Impact of acoustical voice activity detection on spontaneous filled pause classification

Contact Info

Product

Resources

About