| Título: | Eigenvalue Criterion-Based Feature Selectionin Principal Component Analysis of Speech |
| Autores: |
Viszlay, Peter; Department of Electronics and Multimedia Communications
Faculty of Electrical Engineering and Informatics
Technical University of Kosice
Kosice Janecko, Jozef; Department of Electronics and Multimedia Communications Faculty of Electrical Engineering and Informatics Technical University of Kosice Kosice Juhar, Jozef; Department of Electronics and Multimedia Communications Faculty of Electrical Engineering and Informatics Technical University of Kosice Kosice |
| Fecha: | 2012-11-24 |
| Publicador: | Advances in Electrical and Electronic Engineering |
| Fuente: |
 |
| Tipo: |
|
| Tema: | Eigenvalue; feature vector; principal components; selection criterion; variance. |
| Descripción: | This article presents a specific approach for selecting a limited set of most relevant, information rich speech data from the whole amount of training data. The proposed method uses Principal Component Analysis (PCA) to optimally select a lower-dimensional data subset with similar variances. In this paper, three selection algorithms, based on eigenvalue criterion are presented. The first one operates and analyzes the data at the entire speech-recording level. The second one additionally segments each of the recordings into experimentally sized blocks, which theoretically divides a record level into several smaller information richer/poorer blocks. Finally, the third one analyzes all the speech records at the feature vector level. These three approaches represent three different criterion-based selection techniques from the coarsest to the finest data level. The main aim of the presented experiments is to show that PCA trained with the limited subset of data achieves comparable or even better results than PCA trained with the entire speech corpus. In fact, this approach can radically speed up the learning of PCA with much smaller memory and computational costs. All methods are evaluated in Slovak phoneme-based large vocabulary continuous speech recognition task. |
| Idioma: | Inglés |
1 The Applications for the Calculation and Optimization of Passive Optical Networks Attenuation Characteristics por Lafata, Pavel,Vodrazka, Jiri | 6 Binary Modulation Formats in Optical Access Networks por Tejkal, Vladimir,Filka, Miloslav,Reichert, Pavel,Sporik, Jan |
2 The Study of Sensory Properties of Quasi-Singlemode Fiber por Koudelka, Petr; Department of Telecommunications
Faculty of Electrical Engineering and Computer Science
VSB-TU Ostrava
17. listopadu 15/2172
708 33 Ostrava-Poruba
Czech Republic,Latal, Jan,Siska, Petr,Hanacek, Frantisek,Skapa, Jan,Vasinek, Vladimir | 7 Optical Fiber Sensor with Distributed Parameters Based on Optical Fiber Reflectometry por Korenko, Branislav,Jasenek, Jozef,Cervenova, Jozefa,Hlavac, Marek |
3 Fiber Optical Sensor for High Temperatures por Hanacek, Frantisek; Site Administrator,Latal, Jan,Koudelka, Petr,Skapa, Jan,Siska, Petr,Vasinek, Vladimir,Hurta, Jan | 8 Future of Electrotechnics: Ferrofluids por Mayer, Daniel |
4 The Soliton Transmissions in Optical Fibers por Bohac, Leos | 9 Analysis of Simulated Reflection Characteristics of Uniform and Apodized Fiber Bragg Gratings por Gemzicky, Erik,Mullerova, Jarmila |
5 Influence of Magnetic Field on Electric Charge Trasport in Holmium Thin Film at Low Temperature por Dudas, Jan,Gabani, Stanislav,Bagi, Jozef,Goscianska, Iwona,Hodulikova, Anna | 10 WDM in Courses of Communication Technologies por Filka, Miloslav,Bernkopf, Miroslav |