Dissertation > Excellent graduate degree dissertation topics show

Speech Separation Based on Computational Auditory Scene Analysis

Author: LiuJiFang
Tutor: ZhangLei
School: Harbin Engineering University
Course: Signal and Information Processing
Keywords: Speech Separation Computational Auditory Scene Analysis The starting and ending time of the sound
CLC: TP391.42
Type: Master's thesis
Year: 2009
Downloads: 76
Quote: 2
Read: Download Dissertation

Abstract


Computational Auditory Scene Analysis is a knowledge of of auditory psychology and physiology , and let the computer to simulate the human auditory system processes sound . The mixed speech separation based on computational auditory scene analysis is the focus of the present study , the technology for speech recognition , multimedia retrieval, the development of artificial intelligence have important significance . Computational Auditory Scene Analysis , the most important problem is to choose the suitable sound separation clue , most of the existing computational auditory scene analysis system can only be separated with a periodic structure dullness to the neglect of the separation of the voiceless . The study showed the time of the beginning of the sound (onset) and end (offset) by a sudden change of the signal energy to separate the sound , has nothing to do with the voicing of the speech . This article , based on the theory of computational auditory scene analysis , put forward the idea of onset / offset clues to complete the separation of mixed speech . The method adopted in the time domain and frequency domain detection and matching onset / offset, onset / offset on the use of time-frequency information merging fragments to achieve the separation of mixed speech . Can be drawn through the experimental onset / offset clues able to deal with the separation of the voiceless and voiced the same time , solve the problem of treatment with different sound cues voiceless and voiced , making the algorithm more simple ; and onset / offset clue is no limit to the type of mixed sound , whether for music, voice or the noise can be carried out the separation process and obtain a good separation effect .

Related Dissertations

  1. The Research of Speech Segmentation Based on Oscillatory Neural Network,TN912.3
  2. Research on Polyphonic Music Pitch Estimation,TN912.3
  3. Auditory scene analysis and sound source localization algorithm,TN912.3
  4. The Research of Speech Blind Separation Algorithm,TP301.6
  5. Research on Sparse-Decomposition-Based Single-Channel Speech Separation Algorithm,TN912.3
  6. Improved Fast Independent Component Analysis and Its Application on Blind Speech Separation,TN912.3
  7. The Underdetermined Blind Source Separation of Speech Signals,TN912.3
  8. Research on Underdetermined Convolutive Speech Signal Separation Methods,TN912.3
  9. Speech Blind Separation Algorithm,TN912.3
  10. Noisy aliasing speech signals separation method,TN912.3
  11. Study on Speech Enhancement and Separation,TN912.3
  12. Study of Speech Enhancement and Blind Separation Algorithm with Multi-input,TN912.3
  13. On Blind Speech Separation Based on Short-Time Synthesis Overlap-add and It’s DSP Implement,TN912.3
  14. Research on Speech Separation Technology of Mobile Robot Hearing for Navigation System,TN912.3
  15. Research on Speech Separation Based on Spatial Filtering,TN912.3
  16. Underdetermined Anechoic Blind Speech Separation Based on Time-Frequency Sparsity,TN912.3
  17. Research on Blind Separation of Speech Signal with Muti-channel,TN912.3
  18. Separation of Overlapping Speech Based on Computational Auditory Scene Analysis,TN912.3
  19. Complex environment speech enhancement method,TN912.35
  20. Speech Enhancement and Separation Based on Microphone Array,TN912.3

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices > Voice recognition device
© 2012 www.DissertationTopic.Net  Mobile