In speech recognition, discriminative training has proved to be an effective method to improve recognition accuracy. It has successfully improved systems of different scales and different languages. While discriminative training has been developing for over 20 years, it continues to draw attention to researchers and remains to be one of the most important topics in speech recognition to date. D...

It is the problem of computer science that how we detect the handwritten character and word, so it is also the problem in the field of image processing and pattern recognition of computer science. The meaning of handwritten character and word recognition refers to the identification of the characters or word which is written by a human being. Our approach is this, how this problem solved correc...

In part of speech tagging by Hidden Markov Model, a statistical model is used to assign grammatical categories to words in a text. Early work in the field relied on a corpus which had been tagged by a human annotator to train the model. More recently, Cutt ing et al. (1992) suggest that training can be achieved with a minimal lexicon and a limited amount of a priori information about probabilit...

A result of Roquette 3] states that if D is an absolutely irreducible representation of a p-group G over the eld of complex numbers, then D can be realized in K((g) j g 2 G), where is the character of D and K = Q or K = Q(i) according to whether p 6 = 2 or p = 2. Based on Baum and Clausen's 1] algorithm for computing the irreducible representations of supersolvable groups, we give an elementary...

Audio classification has applications in a variety of contexts, such as automatic sound analysis, supervised audio segmentation and in audio information search and retrieval. Extended Baum-Welch (EBW) transformations are most commonly used as a discriminative technique for estimating parameters of Gaussian mixtures, though recently they have been applied in unsupervised audio segmentation. In t...

Video affective content analysis is a fascinating but seldom addressed field in entertainment computing research communities. To recognize affective content in video, a video affective content representation and recognition framework based on Video Affective Tree (VAT) and Hidden Markov Models (HMMs) was proposed. The proposed video affective content recognizer has good potential to recognize t...

We present two efficient search algorithms for real-time spoken language systems. The first called the Word-Dependent N-Best algorithm is an improved algorithm for finding the top N sentence hypotheses. The new algorithm is shown to perform as well as the Exact Sentence-Dependent algorithm presented previously but with an order of magnitude less computation. The second algorithm is a fast match...

We present a simplified derivation of the extended Baum-Welch procedure, which shows that it can be used for Maximum Mutual Information (MMI) of a large class of continuous emission density hidden Markov models (HMMs). We use the extended Baum-Welch procedure for discriminative estimation of MLLR-type speaker adaptation transformations. The resulting adaptation procedure, termed Conditional Max...

* Address all correspondence to Peter W. Roberts, Goizueta Business School, Emory University, 1300 Clifton Road, Atlanta, GA, 30322-2710; (404) 727-8585; [email protected]. Both authors thank Joel Baum, Marc Busch, Heather Haveman, Greta Hsu, Paul Ingram, Ujwal Kayande, Atul Nerkar, Toby Stuart, Harrison White, Mark Zbaracki and Ezra Zuckerman for valuable comments that led to improve...

The paper exposes a non-parametric approach to latent and co-latent modeling of bivariate data, based upon alternating minimization of the Kullback-Leibler divergence (EM algorithm) for complete log-linear models. For categorical data, the iterative algorithm generates a soft clustering of both rows and columns of the contingency table. Well-known results are systematically revisited, and some ...