Dissertation > Excellent graduate degree dissertation topics show

Research on an Embedded Speech Command Processing and Recognition Method

Author: GuoYongLiang
Tutor: ShiYuXia
School: Nanjing University of Aeronautics and Astronautics
Course: Measuring Technology and Instruments
Keywords: Voice command recognition Speech Enhancement Independent Component Analysis Power Spectral Entropy Endpoint detection Wavelet Neural Network Qt Linux
CLC: TN912.3
Type: Master's thesis
Year: 2010
Downloads: 105
Quote: 0
Read: Download Dissertation

Abstract


Speech recognition technology as a convenient, efficient human-computer interaction interface has been applied to the development of handheld devices , voice navigation, intelligent Voice robots , industrial control and other fields . Generation of embedded technology has accelerated the pace of voice command recognition technology from the laboratory to the market , making speech recognition technology to penetrate into all aspects of life , such as voice command recognition technology in the smart home . Therefore, the the embedded voice command to deal with the study of the identification method has important practical significance and broad application prospects . Embedded speech command recognition technology research , in order to improve the effectiveness and robustness of the voice command recognition system for optimized embedded speech command recognition technology in the front - end pretreatment and identification methods and other key issues to expand the research . (1) for the embedded voice command recognition process noise , studied voice enhancement and denoising techniques based on fast independent component analysis , iterative process to improve the core of the algorithm , and gives an improved fast independent component analysis methods . Simulation results show that the new method to accelerate the speed of convergence of the algorithm , the number of iterations , an average reduction of about 30% . (2) for the poor adaptability of the traditional endpoint detection and the detection performance in low signal-to-noise ratio steep drop , a new endpoint detection method . The characteristic parameters of the power spectrum entropy and energy combined endpoint detection . The simulation experiments show that : in stationary noise ( white noise ) , a new method of detection accuracy rate than the endpoint detection method based on spectral entropy about 7% ; non-stationary noise ( pink noise) increased about 5% . (3) The application of wavelet neural network recognition of voice commands , and gives an improved network training method . Experimental results show that the improved method make full use of the sample information to reduce the training time , improve the system recognition rate . (4) Linux system in embedded speech command recognition system and the relaxation starting and ending points of the dynamic time warping techniques , the use of cross-platform graphical interface development tools Qt voice command recognition system for the home environment .

Related Dissertations

  1. Research on Software of TFT-LCD Testing Equipment Based on ARM,TN873.93
  2. Signal Acquisition and Processing Platform Design and Implementation of Trawl Sonar System,S951.2
  3. The Design of Embedded Image Transmission Terminal Based on the TCP/IP Protocol,TP368.1
  4. The Design of the Coal Mine’s Safety Evaluation System Based on Wavelet Neural Network,TD79
  5. Borehole imaging device based on embedded systems research,P634.3
  6. IP QoS technology research,TP393.09
  7. Based on the embedded Web technology Research and Implementation of Dynamic Strain Gauge,TP368.1
  8. Design of Ad hoc Network Application Based on Linux Platform,TN929.5
  9. Design and Implementation of Embedded Multi-parameter Intelligent Evironment Monitoring Systems,TP274
  10. Design of Digital Intelligent Terminal System Doorstation,TP368.1
  11. Intelligent Control System of Chlorine Dosing in Waterworks Based on Embedded System,TU991.2
  12. Design of Intelligent Vehicle Terminal Based on ARM,TP368.1
  13. The Design and Implementation of a Multi-function Floor Display System in Fire Monitoring Network,TN873
  14. Research on Key Technologies of Digital Hearing Aids Based on Auditory Masking,TN912.3
  15. Design of Patch Clamp Amplifier System Based on Qt/E,TP368.1
  16. Design and Research of the family - oriented remote health monitoring system,TP311.52
  17. Design and Development for Process Control Software System of Injection Machine,TQ320.52
  18. An Improved Design of Spectrophotometer with Double Beam,TH744.12
  19. The Research of Car Virtual Instrument System with Linux and ARM,U463.7
  20. The Design of Weld Thermal Cycle Testing Portable Device Based on ARM9,TN305
  21. Design and Implementation of Communication System between Double-chips Based on PCI Bus,TN47

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