Filters








9,965 Hits in 6.2 sec

Between-Class Covariance Correction For Linear Discriminant Analysis in Language Recognition

Abhinav Misra, Qian Zhang, Finnian Kelly, John H.L. Hansen
2016 Odyssey 2016  
Linear Discriminant Analysis (LDA) is one of the most widely-used channel compensation techniques in current speaker and language recognition systems.  ...  In this study, we propose a technique of Between-Class Covariance Correction (BCC) to improve language recognition performance.  ...  In this study, we focus on adapting Linear Discriminant Analysis (LDA) based channel compensation to improve overall system performance.  ... 
doi:10.21437/odyssey.2016-10 dblp:conf/odyssey/MisraZKH16 fatcat:pk7cnyvdsnhuzkvqvf46xjj3oe

Linear Discriminant Analysis Using a Generalized Mean of Class Covariances and Its Application to Speech Recognition

M. SAKAI, N. KITAOKA, S. NAKAGAWA
2008 IEICE transactions on information and systems  
PAPER Special Section on Robust Speech Processing in Realistic Environments Linear Discriminant Analysis Using a Generalized Mean of Class Covariances and Its Application to Speech Recognition Makoto  ...  Linear discriminant analysis (LDA) and heteroscedastic extensions, e.g., heteroscedastic linear discriminant analysis (HLDA) or heteroscedastic discriminant analysis (HDA), are popular approaches to reduce  ...  His major interests in research include automatic speech recognition/speech processing, natural language processing, human interface, and artificial intelligence. He is a fellow of IPSJ.  ... 
doi:10.1093/ietisy/e91-d.3.478 fatcat:m7ja3byp65filiv5bymjdzgcgi

Feature Transformation Based on Generalization of Linear Discriminant Analysis [chapter]

Makoto Sakai, Norihide Kitaoka, Seiichi Nakagaw
2008 Speech Recognition  
Linear discriminant analysis (LDA) is widely used to reduce dimensionality and a powerful tool to preserve discriminative information. LDA assumes each class has the same class covariance.  ...  Heteroscedastic linear discriminant analysis (HLDA) could deal with unequal covariances because the maximum likelihood estimation was used to estimate parameters for different Gaussians with unequal covariances  ... 
doi:10.5772/6378 fatcat:vxyppvhlr5edlcfhcyswj3lfaa

Evaluation of Combinational Use of Discriminant Analysis-Based Acoustic Feature Transformation and Discriminative Training

Makoto SAKAI, Norihide KITAOKA, Yuya HATTORI, Seiichi NAKAGAWA, Kazuya TAKEDA
2010 IEICE transactions on information and systems  
To improve speech recognition performance, acoustic feature transformation based on discriminant analysis has been widely used.  ...  For the same purpose, discriminative training of HMMs has also been used. In this letter we investigate the effectiveness of these two techniques and their combination.  ...  Introduction To improve speech recognition performance, feature transformation such as linear discriminant analysis (LDA) [1] and heteroscedastic discriminant analysis (HDA) [2] are widely used to  ... 
doi:10.1587/transinf.e93.d.395 fatcat:bxkztm3l5jcatbufqbz6q5ukfi

Page 1866 of Mathematical Reviews Vol. 49, Issue 5 [page]

1975 Mathematical Reviews  
On the basis of a training sample from each group he estimates the within-class covariance matrix U”, the between-classes covariance matrix U? and the covariance matrix of the mixed class U.  ...  .) | Polach, Jiti 10199 The possibilities of the linear discriminant method for decision.  ... 

Incorporating Acoustic Feature Diversity Into The Linguistic Search Space For Syllable Based Speech Recognition

Rajesh Hegde, Ramya Rasipuram
2008 Zenodo  
Publication in the conference proceedings of EUSIPCO, Lausanne, Switzerland, 2008  ...  (iii) Linear discriminant analysis is a method used for data classification, it maximizes the ratio of between-class variance to the within-class variance thereby guaranteeing maximal separability between  ...  linear discriminant analysis (LDA).  ... 
doi:10.5281/zenodo.41028 fatcat:aadj2s4c3bbirkpjxspdr7qxru

Comparative Study of Devnagari Handwritten Character Recognition Using Different Feature and Classifiers

Umapada Pal, Tetsushi Wakabayashi, Fumitaka Kimura
2009 2009 10th International Conference on Document Analysis and Recognition  
Many approaches have been proposed by the researchers towards handwritten Indian character recognition and many recognition systems for isolated handwritten numerals/characters are available in the literature  ...  To get idea of the recognition results of different classifiers and to provide new benchmark for future research, in this paper a comparative study of Devnagari handwritten character recognition using  ...  The authors hope this benchmark of results will be helpful to the researchers for future work.  ... 
doi:10.1109/icdar.2009.244 dblp:conf/icdar/PalWK09 fatcat:q4ehkas5jjdqrf35cutb4hsjmu

GESTURE RECOGNITION SYSTEM

Satish Kumar Kotha .
2015 International Journal of Research in Engineering and Technology  
In this approach we have used skin detection techniques for detecting the skin threshold regions, Principle Component Analysis (PCA) algorithm and Linear Discriminant Analysis (LDA) for data compressing  ...  This paper presents a novel approach for the gesture recognition system using software.  ...  Fig 2 : 2 Block diagram of PCA algorithm LDA Algorithm: Linear Discriminant Analysis (LDA) and related Fisher's linear discriminant are the methods used in pattern recognition Fig 3 : 3 Block diagram  ... 
doi:10.15623/ijret.2015.0405019 fatcat:huzxscq7bvfljk3y6gsruzlyt4

"PCA, SFS or LDA: What is the best choice for extracting speaker features"

Abdelghani Harrag, Tayeb Mohamadi
2011 International Journal of Computer Applications  
In this paper, we supply a comparative study for best feature extraction method for speaker recognition system.  ...  A Linear Discriminant Analysis (LDA) method is compared to two well-known feature extraction techniques, namely Principal Component Analysis (PCA) and Sequential Forward Search (SFS).  ...  Linear Discriminant Analysis (LDA) Linear Discriminant Analysis [11] can be summarized as a two phase procedure: in phase one, a class-dependent normalization function collects statistical information  ... 
doi:10.5120/1932-2578 fatcat:wmvuetsmand7laqx24bt6x6cgm

Region dependent linear transforms in multilingual speech recognition

Martin Karafiat, Milos Janda, Jan Cernocky, Lukas Burget
2012 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
In today's speech recognition systems, linear or nonlinear transformations are usually applied to post-process speech features forming input to HMM based acoustic models.  ...  In this work, we experiment with three popular transforms: HLDA, MPE-HLDA and Region Dependent Linear Transforms (RDLT), which are trained jointly with the acoustic model to extract maximum of the discriminative  ...  HLDA transformation allows us to reduce feature dimensionality while preserving information important for discrimination between classes.  ... 
doi:10.1109/icassp.2012.6289014 dblp:conf/icassp/KarafiatJCB12 fatcat:pnftppus7fcxjgtoqavokxxsge

12th Century Ancient Tamil Character Recognition From Temple Wall Inscriptions

S. Rajakumar, V. Subbiah Bharathi
2012 i-manager's Journal on Embedded Systems  
Researchers for the recognition of ancient Tamil languages and scripts are comparatively less with other languages, this is a result of the lack of utilities such as Tamil text databases, dictionaries  ...  Recognition of any ancient Tamil characters with respect to any language is complicated, since the ancient Tamil characters differ in written format, intensity, scale, style, and orientation, from person  ...  , Mirror Image Learning and Linear Discriminant Function are used for comparative study.  ... 
doi:10.26634/jes.1.2.1894 fatcat:5tcuoqosczhxxo2ry5y755ubma

Optical Character Recognition of Amharic Documents

Million Meshesha, C V Jawahar
2007 African Journal of Information & Communication Technology  
This paper presents an Optical Character Recognition (OCR) system for converting digitized documents in local languages.  ...  In this paper, we propose a novel feature extraction scheme using principal component and linear discriminant analysis, followed by a decision directed acyclic graph based support vector machine classifier  ...  Linear Discriminant Analysis Linear discriminant analysis (LDA) selects features based entirely upon their discriminatory potential.  ... 
doi:10.5130/ajict.v3i2.543 fatcat:auqiu6tezvbi3guec55s5camdq

Applying Discriminatively Optimized Feature Transform for HMM-based Off-Line Handwriting Recognition

Jin Chen, Bing Zhang, Huaigu Cao, Rohit Prasad, Prem Natarajan
2012 2012 International Conference on Frontiers in Handwriting Recognition  
Traditionally, linear feature transforms such as Principle Component Analysis (PCA), Linear Discriminative Analysis (LDA) are commonly used.  ...  RDT is one type of non-linear feature transforms which captures the discriminating power much better than traditional linear ones.  ...  Acknowledgement The authors would like to thank Tim Ng and Krishna Subramanian for helpful discussions.  ... 
doi:10.1109/icfhr.2012.182 dblp:conf/icfhr/ChenZCPN12 fatcat:scafjxe4y5anbi4fnkczjafoiu

Phytoplankton recognition using parametric discriminants

Helen McCall, Isabel Bravo, J.Alistair Lindley, Beatriz Reguera
1996 Journal of Plankton Research  
A comparison was made between the use of linear and quadratic discriminant functions for classifying phytoplankton specimens of the genera Dinophysis and Ceratium by means of a general morphometric function  ...  The class distributions were found to fit quadratic boundaries better than linear boundaries. A nine species quadratic discriminant classified within 95% confidence intervals.  ...  Mr R.Williams was responsible for Plymouth Marine Laboratory involvement in the MAST 2 project and contributed to the photography of the specimens.  ... 
doi:10.1093/plankt/18.3.393 fatcat:kmgycghgzve4tgxtjy55aqk7ki

Large-Vocabulary Continuous Speech Recognition Systems: A Look at Some Recent Advances

George Saon, Jen-Tzung Chien
2012 IEEE Signal Processing Magazine  
The projection is designed such as to maximally separate the phonetic classes in the transformed space. The separation is typically measured by a linear discriminant analysis (LDA) criterion [20] .  ...  The projection is trained such as to enhance the discrimination between correct and incorrect word sequences.  ... 
doi:10.1109/msp.2012.2197156 fatcat:sl3fzg2hz5emrpm6srfuc3n3ye
« Previous Showing results 1 — 15 out of 9,965 results