GMM-Based Maghreb Dialect Identification System
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
While Modern Standard Arabic is the formal spoken and written language of the Arab world; dialects are themajor communication mode for everyday life. Therefore, identifying a speaker’s dialect is critical in theArabic-speaking world for speech processing tasks, such as automatic speech recognition or identification. Inthis paper, we examine two approaches that reduce the Universal Background Model (UBM) in the automaticdialect identification system across the five following Arabic Maghreb dialects: Moroccan, Tunisian, and 3dialects of the western (Oranian), central (Algiersian), and eastern (Constantinian) regions of Algeria. Weapplied our approaches to the Maghreb dialect detection domain that contains a collection of 10-secondutterances and we compared the performance precision gained against the dialect samples from a baselineGMM-UBM system and the ones from our own improved GMM-UBM system that uses a Reduced UBMalgorithm. Our experiments show that our approaches significantly improve identification performance overpurely acoustic features with an identification rate of 80.49%.
Abstract
While Modern Standard Arabic is the formal spoken and written language of the Arab world; dialects are themajor communication mode for everyday life. Therefore, identifying a speaker’s dialect is critical in theArabic-speaking world for speech processing tasks, such as automatic speech recognition or identification. Inthis paper, we examine two approaches that reduce the Universal Background Model (UBM) in the automaticdialect identification system across the five following Arabic Maghreb dialects: Moroccan, Tunisian, and 3dialects of the western (Oranian), central (Algiersian), and eastern (Constantinian) regions of Algeria. Weapplied our approaches to the Maghreb dialect detection domain that contains a collection of 10-secondutterances and we compared the performance precision gained against the dialect samples from a baselineGMM-UBM system and the ones from our own improved GMM-UBM system that uses a Reduced UBMalgorithm. Our experiments show that our approaches significantly improve identification performance overpurely acoustic features with an identification rate of 80.49%.
참고문헌 (20)
- K. Kirchhoff, "Cross-dialectal acoustic data sharing for Arabic speech recognition", Proceedings of IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP'04), pp.765~768 (2004)
- D. Vergyri, "Development of a conversational telephone speech recognizer for Levantine Arabic", Proceedings of Interspeech, pp.1613~1616 (2005)
- L. Nour-Eddine, "Pattern Recognition", pp.303~312 (2012), Springer
- A gentle tutorial of the EM algorithm and its application to parameter estimation for Gaussian mixture and hidden Markov models, "J. A. Bilmes", International Computer Science Institute, ~, pp.TR-97-021 (1998)
- P. A. Torres-Carrasquillo, "Approaches to language identification using Gaussian mixture models and shifted delta cepstral features", Proceedings of the 7th International Conference on Spoken Language Processing (ICSLP) (2002)
- P. A. Torres-Carrasquillo, "Dialect identification using Gaussian mixture models", Proceedings of the Speaker and Language Recognition Workshop (ODYSSEY) (2004)
- E. Wong, "Methods to improve Gaussian mixture model based language identification system", Proceedings of the 7th International Conference on Spoken Language Processing (ICSLP) (2002)
- J. L. Gauvain, "Maximum a posteriori estimation for multivariate Gaussian mixture observations of Markov chains", IEEE Transactions on Speech and Audio Processing, 2(2), pp.291~298 (1994)
- D. A. Reynolds, "Speaker verification using adapted Gaussian mixture models", Digital Signal Processing, 10(1), pp.19~41 (2000)
- V. N. Vapnik, "The Nature of Statistical Learning Theory" (1995), Springer
- I. W. Tsang, "Core vector machines : fast SVM training on very large data sets", Journal of Machine Learning Research, 6, pp.363~392 (2005)
- M. Badoiu, "Optimal core-sets for balls", Computational Geometry, 40(1), pp.14~22 (2008)
- L. Nour-Eddine, "Multi-class support vector machines methodology", Proceedings of the 1st International Congress on Models, Optimization, and Security of Systems (ICMOSS) (2010)
- C. W. Hsu, "A comparison of methods for multiclass support vector machines", IEEE Transactions on Neural Networks, 13(2), pp.415~425 (2002)
- S. Asharaf, "Multiclass core vector machine", Proceedings of the 24th International Conference on Machine Learning (ICML), pp.41~48 (2007)
- Multiclass learning at one-class complexity, "S. Szedmak", School of Electronics and Computer Science, University of Southampton, ~ (2005)
- J. C. Dunn, "A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters", Journal of Cybernetics, 3(3), pp.32~57 (1973)
- J. C. Bezdek, "Pattern Recognition with Fuzzy Objective Function Algorithms" (1981), Plenum Press
- J. McLaughlin, "A study of computation speed-UPS of the GMM-UBM speaker recognition system", Proceedings of the 6th European Conference on Speech Communication and Technology (EUROSPEECH), pp.1215~1218 (1999)
- K. S. R. Murty, "Epoch extraction from speech signals", IEEE Transactions on Audio, Speech, and Language Processing, 16(8), pp.1602~1613 (2008)