Combination of Classifiers Decisions for Multilingual Speaker Identification
2
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
State-of-the-art speaker recognition systems may work better for the English language. However, if the samesystem is used for recognizing those who speak different languages, the systems may yield a poorperformance. In this work, the decisions of a Gaussian mixture model-universal background model (GMMUBM)and a learning vector quantization (LVQ) are combined to improve the recognition performance of amultilingual speaker identification system. The difference between these classifiers is in their modelingtechniques. The former one is based on probabilistic approach and the latter one is based on the fine-tuningof neurons. Since the approaches are different, each modeling technique identifies different sets of speakersfor the same database set. Therefore, the decisions of the classifiers may be used to improve the performance. In this study, multitaper mel-frequency cepstral coefficients (MFCCs) are used as the features and themonolingual and cross-lingual speaker identification studies are conducted using NIST-2003 and our owndatabase. The experimental results show that the combined system improves the performance by nearly 10%compared with that of the individual classifier.
Abstract
State-of-the-art speaker recognition systems may work better for the English language. However, if the samesystem is used for recognizing those who speak different languages, the systems may yield a poorperformance. In this work, the decisions of a Gaussian mixture model-universal background model (GMMUBM)and a learning vector quantization (LVQ) are combined to improve the recognition performance of amultilingual speaker identification system. The difference between these classifiers is in their modelingtechniques. The former one is based on probabilistic approach and the latter one is based on the fine-tuningof neurons. Since the approaches are different, each modeling technique identifies different sets of speakersfor the same database set. Therefore, the decisions of the classifiers may be used to improve the performance. In this study, multitaper mel-frequency cepstral coefficients (MFCCs) are used as the features and themonolingual and cross-lingual speaker identification studies are conducted using NIST-2003 and our owndatabase. The experimental results show that the combined system improves the performance by nearly 10%compared with that of the individual classifier.
참고문헌 (26)
- B. S. Atal, "Automatic recognition of speakers from their voices", Proceedings of the IEEE, 64(4), pp.460~475 (1976)
- D. Reynolds, "Robust text-independent speaker identification using Gaussian mixture speaker models", IEEE Transactions on Speech and Audio Processing, 3(1), pp.72~83 (1995)
- P. H. Arjun, "Speaker recognition in indian languages: a feature based approach", Ph.D, pp.2005
- M. Akbacak, "Language normalization for bilingual speaker recognition systems", Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2007), pp.257~260 (2007)
- G. R. Doddington, "The NIST speaker recognition evaluation–overview, methodology, systems, results, perspective", Speech Communication, 31(2), pp.225~254 (2000)
- D. J. Mashao, "Combining classifier decisions for robust speaker identification", Pattern Recognition, 39(1), pp.147~155 (2006)
- E. Kim, "Combination of multiple classifiers for the customer's purchase behavior prediction", Decision Support Systems, 34(2), pp.167~175 (2003)
- T. K. Ho, "Decision combination in multiple classifier systems", IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(1), pp.66~75 (1994)
- C. C. T. Chen, "Speaker identification using hybrid Karhunen–Loeve transform and Gaussian mixture model approach", Pattern Recognition, 37(5), pp.1073~1075 (2004)
- J. Kittler, "On combining classifiers", IEEE Transactions on Pattern Analysis and Machine Intelligence, 20(3), pp.226~239 (1998)
- H. He, "SSC : a classifier combination method based on signal strength", IEEE Transactions on Neural Networks and Learning Systems, 23(7), pp.1100~1117 (2012)
- S. Z. Boujelbene, "Application of combining classifiers for textindependent speaker identification", Proceedings of the 16th IEEE International Conference on Electronics, Circuits, and Systems (ICECS 2009), pp.723~726 (2009)
- V. Hautamaki, "Sparse classifier fusion for speaker verification", IEEE Transactions on Audio, Speech, and Language Processing, 21(8), pp.1622~1631 (2013)
- B. G. Nagaraja, "Multilingual speaker identification by combining evidence from LPR and multitaper MFCC", Journal of Intelligent Systems, 22(3), pp.241~251 (2013)
- "The NIST Year 2003 speaker recognition evaluation plan"[URL]
- T. Kinnunen, "Low-variance multitaper MFCC features : a case study in robust speaker verification", IEEE Transactions on Audio, Speech, and Language Processing, 20(7), pp.1990~2001 (2012)
- J. R. Deller, "Discrete-Time Processing of Speech Signals" (1993), Institute of Electrical and Electronics Engineers
- K. S. Riedel, "Minimum bias multiple taper spectral estimation", IEEE Transactions on Signal Processing, 43(1), pp.188~195 (1995)
- D. J. Thomson, "Spectrum estimation and harmonic analysis", Proceedings of the IEEE, 70(9), pp.1055~1096 (1982)
- M. Hansson, "A multiple window method for estimation of peaked spectra", IEEE Transactions on Signal Processing, 45(3), pp.778~781 (1997)
- D. A. Reynolds, "Encyclopedia of Biometrics", pp.1349~1352 (2009), Springer
- H. S. Jayanna, "Limited data speaker recognition", Ph.D, pp.2009
- J. P. Campbell Jr, "Testing with the YOHO CD-ROM voice verification corpus", Proceedings of 1995International Conference on Acoustics, Speech, and Signal Processing (ICASSP’95), pp.341~344 (1995)
- T. E. F. Filho, "Learning vector quantization in text-independent automatic speaker recognition", Proceedings of Vth Brazilian Symposium on Neural Networks, pp.135~139 (1998)
- S. S. Haykin, "Neural Networks: A Comprehensive Foundation" (1999), Prentice-Hall
- G. Durou, "Multilingual text-independent speaker identification", Proceedings of the Multi-Lingual Interoperability in Speech Technology (MIST) Workshop (1999)