화자확인에서 일정한 결과를 얻기 위한 빠른 순시 확률비 테스트 방법
Fast Sequential Probability Ratio Test Method to Obtain Consistent Results in Speaker Verification
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
A new version of sequential probability ratio test (SPRT) which has been investigated in utterance-length control is proposed to obtain uniform response results in speaker verification (SV). Although SPRTs can obtain fast responses in SV tests, differences in the performance may occur depending on the compositions of consonants and vowels in the sentences used. In this paper, a fast sequential probability ratio test (FSPRT) method that shows consistent performances at all times regardless of the compositions of vocalized sentences for SV will be proposed. In generating frames, the FSPRT will first conduct SV test processes with only generated frames without any overlapping and if the results do not satisfy discrimination criteria, the FSPRT will sequentially use frames applied with overlapping. With the progress of processes as such, the test will not be affected by the compositions of sentences for SV and thus fast response outcomes and even consistent performances can be obtained. Experimental results show that the FSPRT has better performance to the SPRT method while requiring less complexity with equal error rates (EER).
Abstract
A new version of sequential probability ratio test (SPRT) which has been investigated in utterance-length control is proposed to obtain uniform response results in speaker verification (SV). Although SPRTs can obtain fast responses in SV tests, differences in the performance may occur depending on the compositions of consonants and vowels in the sentences used. In this paper, a fast sequential probability ratio test (FSPRT) method that shows consistent performances at all times regardless of the compositions of vocalized sentences for SV will be proposed. In generating frames, the FSPRT will first conduct SV test processes with only generated frames without any overlapping and if the results do not satisfy discrimination criteria, the FSPRT will sequentially use frames applied with overlapping. With the progress of processes as such, the test will not be affected by the compositions of sentences for SV and thus fast response outcomes and even consistent performances can be obtained. Experimental results show that the FSPRT has better performance to the SPRT method while requiring less complexity with equal error rates (EER).
참고문헌 (17)
- Evans. N. W. D, "Assessment of speaker verification degradation due to packet loss in the context of wireless mobile devices", COST275 Workshop on Biometrics over the Internet (2002)
- Norgaard A, "Requirements and biometric technology matching: two approaches from European projects", Proc. of IEEE 39th Annual conference on security technology (2005)
- Yang. Y, "Real time user authentication system for PDA", 2004 IEEE International Conference on Systems,Man and Cybernetics(ICSMC 2004), pp.439~443 (2004)
- Huang. X, "MiPad: a Multimodel Interaction Prototype", Proc. of IEEE Int. Conf. on Acoustics, Speech, and Signal Processing (ICASSP 2001), pp.9~12 (2001)
- Fukunaga. K, "Introduction to statistical pattern recognition", pp.110~119 (1990), Academic Press
- Therrien. C. W., "Decision Estimation and Classification: An Introduction to Pattern Recognition and related Topics", pp.38~41 (1989), John Wiley & Sons
- Text-independent speaker verification using sequential decision, "Harada. K", Technical Report of IEICE, ~, pp.PRMU97-81 (1997)
- Lund. M.A, "A robust sequential test for text-independent speaker verification", J. Acoust. Soc. Am., 99(1), pp.609~621 (1996)
- Noda. H, "A context-dependent Sequential decision for speaker verification", IEICE Trans. INF. & syst., E82-D(10) (1999)
- Seo. C, "Speaker Verification System Using Extended Sequential Probability Ratio Test in PDA", LNCS, 5630, pp.188~193 (2009)
- Noda. H, "Adaptive speaker identification using Sequential Probability Ratio Test", 15th International Conference on Pattern Recognition, 통권 : 3, pp.262~265 (2000)
- Reynolds. D, "Robust text-independent speaker identification using Gaussian mixture speaker models", IEEE Trans. on SAP, 3(1), pp.72~82 (1995)
- 서창우, "화자확인에서 특징벡터의 순시 정보와 선형 변환의 효과적인 적용", 말소리와 음성과학, 1(4), pp.127~132 (2009), 한국음성학회
- Young. S, "The HTK Book (for HTK version 3.2)" (2002), Cambridge University Engineering Department
- Parthasarathy. S, "General phrase speaker verification using sub-word background models and likelihood ratio scoring", ICSLP-96 (1996)
- Matsui. T, "Likelihood normalization for speaker verification using a phoneme and speaker-independent model", Speech Communication, 17, pp.109~116 (1995)
- Lind. Y, "An algorithm for vector quantizer design", IEEE Trans. Commun., 28, pp.84~95 (1980)