히스토그램 변환에서 기준분포의 표준편차 변경에 따른 강인한 화자인증 성능 개선
Performance Improvement of Robust Speaker Verification According to Various Standard Deviations of a Reference Distribution in Histogram Transformation
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Additive noise and channel mismatch strongly degrade the performance of speaker verification systems, as they distort the features of speech. In this paper a histogram transformation technique is presented to improve the robustness of text-independent speaker verification systems. The technique transforms the features extracted from speech such that their histogram is conformed to a reference distribution. The effect of different standard deviations for the reference distribution is investigated. Experimental results indicate that, in channel mismatched environments, the proposed technique offers significant improvements over existing techniques. We also verify performance improvement of the proposed method using statistics.
Abstract
Additive noise and channel mismatch strongly degrade the performance of speaker verification systems, as they distort the features of speech. In this paper a histogram transformation technique is presented to improve the robustness of text-independent speaker verification systems. The technique transforms the features extracted from speech such that their histogram is conformed to a reference distribution. The effect of different standard deviations for the reference distribution is investigated. Experimental results indicate that, in channel mismatched environments, the proposed technique offers significant improvements over existing techniques. We also verify performance improvement of the proposed method using statistics.
참고문헌 (14)
- 유하진, "화자인식 기술 및 국내외시장 동향", 대한음성학회 2004 봄 학술대회 발표논문집, pp.91~97 (2004)
- Auckenthaler, R., "Score normalization for text-independent speaker verification systems", Digital Signal Processing, 10, pp.42~54 (2000)
- Auckenthaler, R., "Improving a GMM speaker verification system by phonetic weighting", Proc. International Conf. on Acoustics Speech Signal Proc., 통권 : 1, pp.313~316 (1999)
- Reynolds, D.A., "Speaker verification using adapted Gaussian mixture models", Digital Signal Processing, 10, pp.19~41 (2000)
- Campbell, J.P., "Speaker recognition: a tutorial", Proceedings of the IEEE, 85(9), pp.1437~1462 (1997)
- de la Torre, A., "Histogram equalization of speech representation for robust speech recognition", IEEE Transactions on Speech and Audio Processing, 13(3), pp.355~366 (2005)
- Skosan, M., "Modified segmental histogram equalization for robust speaker verification", Pattern Recognition Letters, 27(5), pp.479~486 (2006)
- Pelecanos, J., "Feature warping for robust speaker verification", Proc. Odyssey 2001, 통권 : 1, pp.213~218 (2001)
- Xiang, B., "Short-time Gaussianization for robust speaker verification", Proc. International Conf. on Acoustics Speech Signal Proc., 통권 : 1, pp.681~684 (2002)
- 서영주, "클래스 히스토그램 등화 기법에 의한 강인한 음성인식", 말소리, 1(60), pp.145~164 (2006), 대한음성학회
- Grundland, M., "Color histogram specification by histogram warping", Proc. SPIE, 통권 : 5667, pp.610~621 (2005)
- Young, S., "The HTK Book" (2001), Cambridge University Engineering Department
- 성태제, "현대 기초통계학의 이해와 적용" (2007), 교육과학사
- Reynolds, D., "Speaker identification and verification using Gaussian mixture speaker models", Speech Communication, 17(1), pp.91~108 (1995)