Dissertation > Excellent graduate degree dissertation topics show

Research on the Comparison and Parallelization of Discriminative Training of Acoustic Model

Author: HeXueLiang
Tutor: HuangYunQing
School: Xiangtan University
Course: Computational Mathematics
Keywords: Automatic Speech Recognition HMM Acoustic Model DiscriminativeTraining Parallel Algorithm HTK
CLC: TN912.34
Type: Master's thesis
Year: 2013
Downloads: 9
Quote: 0
Read: Download Dissertation

Abstract


Research on the Automatic Speech Recognition has been rapid development inrecent years. More and more voice products are constantly applied in our daily life,all of which are based on the advance of voice technology, especially the research ofthe discriminative training. It can efectively improve the recognition performanceof the system when it is applied to voice continuous speech recognition system. Ithas become a new hotspot in the ASR field in recent years.This article first introduces the development history of speech recognition andthe main stages in the process of development, and then introduces the principalcompositions of the speech recognition system, which we show the detailed intro-duction.Secondly, we introduce the acoustic model based on HMM, and the three basicquestions need to be solved when applied to the speech recognition. We show howto use some algorithms to solve the relevant problems of the speech recognition.Thirdly, we introduce the methods of the discriminative training which is com-monly. At first, we introduced the Maximum Likelihood Estimation (MLE) method,and through the Bayesian Decision Rule to elicit the discriminative training. Byoptimizing the diferent parts, we can obtain the diferent discriminative trainingmethods, e.g. MMIE, MCE, MPE/MWE, and some other the commonly train-ing algorithms. And then we compare the traits of these discriminative trainingmethods. At last, we operate phonetic experiment on HTK platform.Finally, we introduce the parallelization of the discriminate training, and putforward a new parallel algorithm about it. From the last experiment, we find that,the algorithm can reduce the training time efectively without afecting the recogni-tion results when it is combined with discriminate training.

Related Dissertations

  1. Visual Feedback and Memory Behavior Based GPU Parallel Ant Colony Algorithm,TP301.6
  2. Design and Implementation of Image Authentication on CUDA Platform,TP391.41
  3. A Clinical Comparative Study on Myocardial Protective Effect between HTK Solution and Cold Blood Cardioplegia in Infants with CHD Undergoing Open Heart Surgery,R726.5
  4. Research on high-density signal de-noising method based on Radon transform point earthquake,P631.4
  5. Research on Parallel Algorithms of the Fast Algorithm for Spherical Harmonic Expansions,TP301.6
  6. Research on Hmm-based Speech Recognition System of the Robot,TN912.34
  7. The Technology Research of Endpoint Detection and Keywords Detection of Speech,TN912.3
  8. Design and Implementation of Fault-Tolerant Parallel Algorithm for On-board Computer,TP302.8
  9. Mobile robot voice recognition control simulation system design and implementation,TN912.34
  10. Research on Tree Structure Based Power Flow Parallel Calculation of Power System,TM744
  11. Design and Implementation of Chinese Continuous Speech Recognition System Based on HTK,TN912.34
  12. Research on Dual-DSP Board and Realization of Multi-beam Bathymetry Detection Algorithm,U666.7
  13. Parallel Triangle Processing with CUDA,TP391.41
  14. Parallel ant algorithm for the traveling salesman problem,TP301.6
  15. Speech Recognition Algorithm and Implementation,TN912.34
  16. The Protective Effects of STH Solution Combined with HTK Solution on Long Time Ischemic-reperfusion Injury in Isolated Rat,R965
  17. The Effect of Cold Autologous Blood Cardioplegic Solution to Energy Metabolism and Mitochondria of Immature Cardiocyte,R654.2
  18. HTK solution on oxygen-deficient immature myocardial protection in clinical research,R654.2
  19. A Study on Effectiveness of an ASR-based Online Software on Chinese College Students’ Phoneme Learning,H319
  20. Design of Speech Control System on Av Bimodal Information and Its Realization,TN912.2
  21. Protein Folding Prediction Based on Parallel Hybrid Particle Swarm Optimization Algorithm,TP301.6

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Communicate > Electro-acoustic technology and speech signal processing > Speech Signal Processing > Speech Recognition and equipment
© 2012 www.DissertationTopic.Net  Mobile