ABSTRACT
Speech recognition is one of the next generation technologies for human-computer interaction. Automatic Speech Recognition (ASR) is a technology that allows a computer to recognize the words spoken by a person through telephone, microphone or other devices. The various stages of the speech recognition system are pre-processing, segmentation of speech signal, feature extraction of speech and recognition of word. Among many speech recognition systems, continuous speech recognition system is very important and most popular system. This paper proposes the time-domain features and frequency-domain features based on fuzzy knowledge for continuous speech segmentation task via a nonlinear speech analysis. Short-time Energy and Zero-crossing Rate are time-domain features, and Spectral Centroid is frequency-domain feature that the system will calculate in each point of speech signal in order to exploit relevant information for generating the significant segments. Fuzzy Logic technique will be used not only to fuzzify the calculated features into three complementary sets namely: low, middle, high but also to perform a matching phase using a set of fuzzy rules. The outputs of the Fuzzy Logic are phonemes, syllables and disyllables of Myanmar Language. The result of the system will recognize the continuous words of input speech.
- T. T. Thet, J. Na and W. K. Ko, "Word Segmentation for the Myanmar Language", Journal of Information Science, 2008. Google ScholarDigital Library
- T. M. Tun and K. T. Lynn, "Myanmar Continuous Speech to Isolated Word Segmentation", Engineering and Technology, IJSRSET, Issue 2, Volume 1, 2015.Google Scholar
- Haykin, S (2001), Minimum mean square error adaptive filter. In Adaptive Filter Theory, 4th ed. Prentice Hall, Upper Saddle River, 183--228.Google Scholar
- M. M. Rahman and M. A. Bhuiyan, "Continuous Bangla Speech Segmentation using short time Speech Features Extraction Approaches., IJACSA, Volume 3, No. 11, 2012.Google Scholar
- T. Zhang and J. C. C. Kuo, "Hierarchical classification of audio data for archiving and retrieving", In International Conference on Acoustics, Speech and Signal Processing, volume VI, pages 3001--3004. IEEE, 1999. Google ScholarDigital Library
- L R Rabiner and M R Sambur, "An Algorithm for determining the endpoints of Isolated Utterances", The Bell System Technical Journal, February 1975, pp 298--315.Google Scholar
- T Giannakopoulos, "Study and application of acoustic information for the detection of harmful content and fusion with visual information" Ph.D. dissertation, Dept. of Informatics and Telecommunications, University of Athens, Greece, 2009.Google Scholar
- Vimala, C., Radha, V., A review on speech recognition challenges and approaches", World Computer. Sci. Inf. Technol., 2012, 2, (1), pp. 1--7.Google Scholar
- Mr. Sridhar Chandramohan Iyer, Speaker Recognition System using Coefficients and Correlation Approaches in MATLAB, IJERT, Vol. 3 Issue 5, May 2014.Google Scholar
Index Terms
- Myanmar Continuous Speech Recognition System Using Fuzzy Logic Classification in Speech Segmentation
Recommendations
Fuzzy-based algorithm for Fongbe continuous speech segmentation
Text-independent speech segmentation is a challenging topic in computer-based speech recognition systems. This paper proposes a novel time-domain algorithm based on fuzzy knowledge for continuous speech segmentation task via a nonlinear speech analysis. ...
Continuous Punjabi speech recognition model based on Kaldi ASR toolkit
In this paper, continuous Punjabi speech recognition model is presented using Kaldi toolkit. For speech recognition, the extraction of Mel frequency cepstral coefficients (MFCC) features and perceptual linear prediction (PLP) features were extracted ...
Segment Matrix Vector Quantization and Fuzzy Logic for Isolated-Word Speech Recognition
ISMVL '95: Proceedings of the 25th International Symposium on Multiple-Valued LogicAbstract: A novel speech recognition approach using segment matrix vector quantization (SMVQ) and fuzzy logic recognizer (FLR) is presented. SMVQ incorporates time sequence information and segment characteristics of speech signals. Firstly, the feature ...
Comments