Abstract

Support Vector Machines have recently shown dramatic performance gains in many application areas. We show that the same gains can be realized in the area of speaker recognition via sequence kernels. A sequence kernel provides a numerical comparison of speech utterances as entire sequences rather than a probability at the frame level. We introduce a novel sequence kernel derived from generalized linear discriminants. The kernel has several advantages. First, the kernel uses an explicit expansion into "feature space"--this property allows all of the support vectors to be collapsed into a single vector creating a small speaker model. Second, the kernel retains the computational advantage of generalized linear discriminants trained using mean-squared error training. Finally, the kernel shows dramatic reductions in equal error rates over standard mean-squared error training in matched and mismatched conditions on a NIST speaker recognition task.

Citations

3909	Neural Networks for Pattern Recognition - Bishop - 1995
315	Exploiting generative models in discriminative classifiers - Jaakkola, Haussler - 1998
190	Bengio S: SVMTorch: Support Vector Machines for Large-Scale Regression Problems - Collobert
78	Probabilities for SV machines - Platt - 1999
28	recognition using neural networks and conventional classifiers - FARRELL, MAMMONE, et al. - 1994
26	Pattern Classification - Schürmann - 1996
22	Automatic Speaker Recognition Using Gaussian Mixture Speaker Models - Reynolds - 1995
21	Hybrid SVM/HMM architectures for speech recognition - Ganapathiraju, Hamaker, et al. - 2000
17	Assaleh, “Polynomial classifier techniques for speaker verification - Campbell, T - 1999
11	The NIST speaker recognition evaluation—Overview, methodology, systems, results, perspective - Doddington, Przybocki, et al. - 2000
10	A speaker verification system using alphanets - Carey, Parris, et al. - 1991
4	Mahesan Niranjan, “Data-dependent kernels in SVM classification of speech patterns - Smith, Gales - 2001
4	A computationally scalable speaker recognition system - Campbell, Broun - 2000
3	Support vector machines for verification and identification - Wan, Campbell
3	Rabiner and Biing-Hwang - Lawrence - 1993
2	A hybrid GMM/SVM approach to speaker recognition - Fine, Navrátil, et al. - 2001