Sparsity Analysis and Compensation for i-Vector Based Speaker Verification
Over recent years, i-vector based framework has been proven to provide state-of-art performance in speaker verification. Most of the researches focus on compensating the channel variability of i-vector. In this paper we will give an analysis that in the case that the duration of enrollment or test utterance is limited, i-vector based system may suffer from biased estimation problem. In order to solve this problem, we propose an improved i-vector extraction algorithm which we term Adapted First order Baum-Welch Statistics Analysis (AFSA). This new algorithm suppresses and compensates the deviation of first order Baum-Welch statistics caused by phonetic sparsity and phonetic imbalance. Experiments were performed based on NIST 2008 SRE data sets, Experimental results show that 10 %–15 % relative improvement is achieved compared to the baseline of traditional i-vector based system.
KeywordsSpeaker verification i-vector Phonetic sparsity Adapted first order Baum-Welch statistics analysis (AFSA)
This article was supported by the National Natural Science Foundation of China (NSFC) under Grants No. 61271349, 61371147 and 11433002.
- 1.Bonastre, J.F., Scheffer, N., Matrouf, D., Fredouille, C., Larcher, A., Preti, A., Pouchoulin, G., Evans, N.W., Fauve, B.G., Mason, J.S.: Alize/spkdet: a state-of-the-art open source software for speaker recognition. In: Odyssey, p. 20 (2008)Google Scholar
- 2.Bousquet, P.M., Larcher, A., Matrouf, D., Bonastre, J.F., Plchot, O.: Variance-spectra based normalization for i-vector standard and probabilistic linear discriminant analysis. In: Speaker and Language Recognition Workshop (IEEE Odyssey) (2012)Google Scholar
- 3.Bousquet, P.M., Matrouf, D., Bonastre, J.F.: Intersession compensation and scoring methods in the i-vectors space for speaker recognition. In: INTERSPEECH, pp. 485–488 (2011)Google Scholar
- 5.Kenny, P.: Joint factor analysis of speaker and session variability: Theory and algorithms. CRIM, Montreal, (Report) CRIM-06/08-13 (2005)Google Scholar
- 6.Kenny, P.: Bayesian speaker verification with heavy-tailed priors. In: Odyssey, p. 14 (2010)Google Scholar
- 10.Pelecanos, J., Sridharan, S.: Feature warping for robust speaker verification (2001)Google Scholar