1 / 15

MINUET Musical Interference Unmixing Estimation Technique

MINUET Musical Interference Unmixing Estimation Technique. Scott Rickard, Conor Fearon Department of Electronic & Electrical Engineering University College Dublin, Ireland Radu Balan, Justinian Rosca Siemens Corporate Research, Princeton, NJ. 18 th March 2004. CISS04. MINUET: The Problem.

melody
Download Presentation

MINUET Musical Interference Unmixing Estimation Technique

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. MINUETMusical Interference Unmixing Estimation Technique Scott Rickard, Conor Fearon Department of Electronic & Electrical Engineering University College Dublin, Ireland Radu Balan, Justinian Rosca Siemens Corporate Research, Princeton, NJ. 18th March 2004 CISS04

  2. MINUET: The Problem • Given x and n’ • Find s

  3. Classical Solution(Adaptive Filtering)

  4. Adaptive Algorithms • Least-Mean Square (LMS) Algorithm - minimises mean-square error • Recursive Least Squares (RLS) Algorithm - minimises sum of squares of error

  5. Problem! • Performance drastically deteriorates with small phase and synchronisation errors. • Mixture: • No error: • Delayed by 1 sample: • Delayed by 10 samples:

  6. W-Disjoint Orthogonality • At every point in the t-f representation of a mixture, only one source is active.

  7. MINUET Solution • Consider simple problem: • Create Mask: • Solution:

  8. Synchronisation Errors? • The performance of time-frequency masking with respect to small phase and synchronisation errors is extremely robust. • Mixture: • No error: • Delayed by 1 sample: • Delayed by 10 samples:

  9. SNR improvement

  10. Performance Measures • SNR is a standard performance measure • But what about speech quality? • Incorrect partitioning of t-f domain reduces intelligibility of output. • Introduce measure of WDO: O. Yilmaz and S. Rickard, "Blind Separation of Speech Mixtures via Time-Frequency Masking", IEEE Transactions on Signal Processing, To appear, July 2004.

  11. WDO

  12. MINUET Channel Estimate • Find set of t-f points, S, such that for

  13. Adaptive Testing Unity Channel: Random Channel:

  14. Conclusions and Future Work • MINUET estimates the channel and removes interference using instantaneous t-f magnitudes only. • This creates extraordinary robustness to phase errors when compared to classical adaptive filtering methods. • Improvements in t-f masking still necessary to increase intelligibility. • Algorithm complexity has not yet been considered. • We presented pilot tests serving as proof of concept only. • More realistic testing must be done to genuinely assess performance. • MINUET will be effective for any signals which are WDO.

  15. Thank you for your attention!

More Related