Loris for Your Cough

Roshan Mansinghani | Esmeralda Martinez | James McDougall | Travis McPhail
Home
Background
McAulay-Quatieri Method
Bandwidth Enhancement
Algorithm
   Loris Application
   Our Algorithm
Results
Conclusion
Poster
Group

Conclusion

 

While we removed the cough, we lost quality in the recorded sound. Using a different alogorithm, but the same Bandwidth Enhancement analysis, we could improve on our results.

Possible improvements:

  • Increase the number of harmonics used to calculate the closeness of partials in the "outCrowd" to the partials in the "inCrowd"
  • Match Filter approach
  • Create templates of different types of noise: claps, coughs, etc.
  • After the partials have been created.  Perform a match filter operation on them to see how close they are to a given "noise template"  if they differ greatly from the "noise template" add and are relatively close to a harmonic of a dominating partial (one of those in the "inCrowd" then add them to the "inCrowd")