Masafumi Nishida
ABSTRACT
Speech feature variations are mainly attributed to variations in phonetic and speaker information included in speech data. If these two types of information are separated from each other, more robust speaker clustering can be achieved. Principal component analysis transformation can separate speaker information from phonetic information, under the assumption that a space with large within-speaker variance is a "phonetic subspace" and a space within-speaker variance is a "phonetic sub-space". We propose a speaker clustering method based on non-negative matrix factorization using a Gaussian mixture model trained in the speaker subspace. We carried out comparative experiments of the proposed method with conventional methods based on Bayesian information criterion and Gaussian mixture model in an observation space. The experimental results showed that the proposed method can achieve higher clustering accuracy than conventional methods.
- S. E. Tranter and D. A. Reynolds, "An Overview of Automatic Speaker Diarization Systems", IEEE Transactions on AudioGoogle Scholar
- D. A. Reynolds and P. Torres-Carrasquillo, "Approaches and Applications of Audio Diarization", Proc. ICASSP, Vol.5, pp. 953--956, 2005. Google ScholarCross Ref
- S. Chen and P. Gopalakrishnan, "Speaker Environment and Channel Change Detection and Clustering via the Bayesian Information Criterion", Proc. DARPA Broadcast News Transcription and Understanding Workshop, pp. 127--132, 1998.Google Scholar
- S. Cheng, H. Wang, H. Fu, "BIC-based Speaker Segmentation Using Divide-and-conquer Strategies with Application to Speaker Diarization", IEEE Transactions, Vol.18, pp. 141--157, 2010.Google ScholarDigital Library
- K. Iso, "Speaker Clustering Using Vector Quantization and Spectral Clustering", Proc. ICASSP, pp. 4986--4989, 2010. Google ScholarCross Ref
- M. Nishida and T. Kawahara, "Speaker Model Selection Based on the Bayesian Information Criterion Applied to Unsupervised Speaker Indexing", IEEE Transactions on Speech and Audio Processing, Vol.13, No.4, pp. 583--592, 2005. Google ScholarCross Ref
- D. A. Reynolds, E. Singer, B. A. Carlson, G. C. O'Leary, J. J. McLaughlin, and M. A. Zissman, "Blind Clustering of Speech Utterances based on Speaker and Language Characteristics", Proc. ICSLP, pp. 3193--3196, 1998.Google Scholar
- L. Viet Bac, O. Mella, and D. Fohr, "Speaker Diarization using Normalized Cross Likelihood Ratio", Proc. Interspeech, pp. 1869--1872, 2007.Google Scholar
- M. Nishida and Y. Ariki, "Speaker Recognition by Separating Phonetic Space and Speaker Space", Proc. EUROSPEECH, Vol. 2, pp. 1381--1384, 2001.Google Scholar
- M. Nishida and S. Yamamoto, "Speaker Clustering Based on Non-negative Matrix Factorization", Proc. Interspeech, pp. 949--952, 2011.Google Scholar
- Sneath, P. H. A. and Sokal, R. R, "Numerical Taxonomy", W. H. Freeman and Company, 1973.Google Scholar
- K. Maekawa, "Corpus of Spontaneous Japanese: Its Design and Evaluation", Proc. ISCA & IEEE Workshop on SSPR, pp. 7--12, 2003.Google Scholar
Index Terms
- Speaker Clustering Based on Non-Negative Matrix Factorization Using Gaussian Mixture Model in Complementary Subspace
Recommendations
Rapid speaker adaptation in latent speaker space with non-negative matrix factorization
A novel speaker adaptation algorithm based on Gaussian mixture weight adaptation is described. A small number of latent speaker vectors are estimated with non-negative matrix factorization (NMF). These latent vectors encode the distinctive systematic ...
Speaker Verification Using Adapted Bounded Gaussian Mixture Model
2018 IEEE International Conference on Information Reuse and Integration (IRI)In this paper, we propose the application of bounded Gaussian mixture model (BGMM) to speaker verification. In the proposed approach, BGMM is employed for universal background model (UBM) and adapted speaker model. The proposed UBM is a large BGMM trained ...
Speaker Verification Using Adapted Gaussian Mixture Models
Reynolds, Douglas A., Quatieri, Thomas F., and Dunn, Robert B., Speaker Verification Using Adapted Gaussian Mixture Models, Digital Signal Processing10(2000), 19 41.In this paper we describe the major elements of MIT Lincoln Laboratory's Gaussian ...
Comments