Dissertation > Excellent graduate degree dissertation topics show

Research on Music Retrieval Based on Content

Author: HuaHui
Tutor: LiuChuanCai
School: Nanjing University of Technology and Engineering
Course: Applied Computer Technology
Keywords: Content - based music retrieval Endpoint detection DTW algorithm Bayesian decision Melody matching algorithm
CLC: TP391.3
Type: Master's thesis
Year: 2008
Downloads: 151
Quote: 0
Read: Download Dissertation

Abstract


Humming - based music retrieval system , the user simply hum the melody sing a song fragment into the microphone , you can get the music you want to retrieve . In the field of pattern recognition , music retrieval algorithm to get more and more attention because of its easy and efficient features , which humming sound endpoint detection , feature extraction , the music melody matching , and processing of environmental noise is music retrieval the difficulty . This paper mainly around the content - based music retrieval to carry out research work as follows : ( 1) Feature Extraction : analysis of the audio signal of the characteristics of the time domain and frequency domain characteristics , and different combinations of features to be applied to different types of audio retrieval . ( 2) endpoint detection : the traditional endpoint detection method based on endpoint detection algorithm based on Bayesian decision , but also to achieve a zero rate of endpoint detection algorithm based energy tracking endpoint detection algorithms, and center cut Bohou Ji . Experiments to compare the three algorithms to verify the validity of the endpoint detection algorithm based on Bayesian decision , to achieve a better note segmentation . (3) The pitch extraction : In traditional endpoint detection method based on the use of the autocorrelation function of the pitch period of the voiced speech segment extracted . (4) the melody matching algorithm : discourse analysis based on the format of the music data classification , respectively characteristic the melody feature matching algorithm by the DTW algorithm melody match . The music signal by the endpoint detection note segmentation, to improve the accuracy of the notes feature extraction . On the basis of the endpoint detection , improved DTW, significantly improves the recognition results .

Related Dissertations

  1. Power spectrum estimation in the broadband ADCP Signal Detection Research and Application,TN911.23
  2. Telephone-based channel voiceprint recognition algorithm,TN912.34
  3. MFCC -based speech recognition system to improve research and design,TN912.34
  4. Research on Endpoint Detection Algorithm and Implementation in Hardware,TN912.34
  5. Endpoint Detection of Speech Signal Based on Empirical Mode Decomposition,TN912.3
  6. DFI Research in P2P Traffic Identification,TP393.06
  7. Face Recognition Based on Bayesian Decision Fusion of the Algorithm,TP391.41
  8. Research and Implementation of Adaptive Low-Rate Speech Coding,TN912.3
  9. Research on Robot Path Planning Based on Bayes Decision and Ant Colony Algorithm in a Complex Environmnet,TP242
  10. Research on Text-independence of Open-set Speaker Recognition,TN912.34
  11. Research of the Speech Recognition Technology Based on DSP,TN912.34
  12. Research and Implementation of Speech Recognition System for Mobile Robot,TN912.34
  13. Research on Image Denoising Algorithm Based on the Minimum Error Rate Bayesian Decision-making and Smoothing Filtering,TP391.41
  14. The Research and Implement of Isolated Word Speech Recognition Algorithm Based on DTW Model,TN912.34
  15. Speaker Recognition System and DSP Implementation,TN912.34
  16. ASIC Design of Specific Speaker Isolated Speech Recognition System,TN912.34
  17. The Optimization of Probabilistic Neural Network Structure and Its Applications,TP183
  18. Microphone Array Speech Enhancement Based on Filterbanks Generalized Sidelobe Canceller,TN641
  19. Audit System of Enterprise Site Service Based on Content Filtering Technology,TP393.08
  20. The Cough Signal Spotting in Continuous Speech,TN912.34
  21. 3D face lip animation voice-driven research,TN912.34

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Retrieval machine
© 2012 www.DissertationTopic.Net  Mobile