Mixture of Support Vector Machines for HMM based Speech Recognition

Krüger, Sven E.; Schafföner, Martin; Katz, Marcel; Andelic, Edin; Wendemuth, Andreas

doi:10.1109/icpr.2006.804

Cited by 15 publications

(4 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These works can be divided into two branches: mixture of hyperplanes and mixture of hyperspheres. The works presented in [6,12,14] applied divide-and-conquer strategy to partition the input space into many disjoint regions, and in each region, a linear or nonlinear SVM can be employed to classify data. In [19], multiple hyperplanes were used for learning to rank.…”

Section: Related Workmentioning

confidence: 99%

Mixture of hyperspheres for novelty detection

Nguyen¹,

Lai²,

Nguyen³

et al. 2016

Vietnam J Comput Sci

View full text Add to dashboard Cite

In this paper, we present a mixture of support vector data descriptions (mSVDD) for one-class classification or novelty detection. A mixture of optimal hyperspheres is automatically discovered to characterize data. The model includes two parts: log likelihood to control the fit of data to model (i.e., empirical risk) and regularization quantizer to control the generalization ability of model (i.e., general risk). Expectation maximization (EM) principle is employed to train our proposed mSVDD. We demonstrate the advantage of the proposed model: if learning mSVDD in the input space, it simulates learning a single hypersphere in the feature space and the accuracy is thus comparable, but the training time is significantly shorter.

show abstract

Section: Related Workmentioning

confidence: 99%

Mixture of hyperspheres for novelty detection

Nguyen¹,

Lai²,

Nguyen³

et al. 2016

Vietnam J Comput Sci

View full text Add to dashboard Cite

show abstract

“…Competing approaches [24], [7], [23] employ a divide-and-conquer strategy similar in spirit to the mixture of expert framework [20] in the partitioning of the input space into disjoint regions. Once the space has been partitioned, a local non-linear SVM classifier is trained on samples falling in each region.…”

Section: Discussionmentioning

confidence: 99%

Mixing Linear SVMs for Nonlinear Classification

Robles-Kelly

Zhou

2010

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

Abstract-In this paper, we address the problem of combining linear Support Vector Machines (SVMs) for classification of large-scale non-linear data sets. The motivation is to exploit both the efficiency of linear SVMs in learning and prediction and the power of non-linear SVMs in classification. To this end, we develop a linear SVM mixture model that exploits a divide-and-conquer strategy by partitioning the feature space into sub-regions of linearly separable data-points and learning a linear SVM for each of these regions. We do this implicitly by deriving a generative model over the joint data and label distributions. Consequently, we can impose priors on the mixing coefficients and do implicit model selection in a top-down manner during the parameter estimation process. This guarantees the sparsity of the learned model. Experimental results show that the proposed method can achieve the efficiency of linear SVMs in the prediction phase while still providing a classification performance comparable to non-linear SVMs.

show abstract

“…Speech recognition presents a difficult problem. Current research is focused on hidden Markov models (HMMs) and ANNs [9], [11]- [13]. Both the HMM and ANN approaches to speech recognition are complex software-based systems that are slow and require a lot of processing power.…”

Section: Introductionmentioning

confidence: 99%

Implementing template matching logic in a standard flash memory cell

Milacic

Dimitrijev

2013

FACTA U EE

View full text Add to dashboard Cite

Current research into classification methods is almost exclusively software based, resulting in systems that perform well but are invariably slow when faced with large databases. The goal is therefore to create a hardware classification system that is much faster. In this paper, we introduce the concept of template matching logic and propose the use of a standard flash memory cell array to perform bit by bit template matching. The proposed system is based on a novel architecture that is unique and separate from existing architectures that make use of flash memory cell arrays. Verification is achieved by speech recognition simulations on the TIMIT database. Simulations of the system show results of 94.5 % recognition accuracy on clean words and 88.0 % recognition accuracy on test words with a signal-to-noise ratio of 5 dB. The results compare favorably to similar isolated word recognition tasks performed with software based methods.

show abstract

Mixture of Support Vector Machines for HMM based Speech Recognition

Cited by 15 publications

References 7 publications

Mixture of hyperspheres for novelty detection

Mixture of hyperspheres for novelty detection

Mixing Linear SVMs for Nonlinear Classification

Implementing template matching logic in a standard flash memory cell

Contact Info

Product

Resources

About