Timeline for What methodology to use for discrimination of different (musical?) tones
Current License: CC BY-SA 3.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 29, 2012 at 6:35 | comment | added | Spacey | @pichenettes I have thought some more about this, and I still do not understand the following: You said "and the ratio of the larger secondary peak of the autocorrelation to r0 as a pitch confidence indicator. and also Signals of type 2 will have a low σf (steady pitch) and a high σv and σe (variable strength). - can you please elaborate on this? I suppose it comes down to not really understanding why the 'pitch confidence' should be different for signals 1 or 2... | |
| Feb 17, 2012 at 16:50 | comment | added | pichenettes | 6/ I think the best rule to trigger an alarm is to have something "if x% of the last T frames have been classified as 1, ring alarm for event 1". Adjust x and T depending on the required reaction time. | |
| Feb 17, 2012 at 16:49 | comment | added | pichenettes | 5/ Yes, this is the sequence of classifier outputs, which eventually will have to be post-processed if there are spurious misclassifications - you are taking here advantage of the fact that you know that the sequence of values is pretty stable over long stretches | |
| Feb 17, 2012 at 16:49 | comment | added | pichenettes | 3/ Take a STFT "slice", multiply it element-wise by its conjugate, take the inverse Fourier transform and you have the autocorrelation of the time domain window. 4/ Correct, your 3d point are the three standard deviations over the 101 frames. | |
| Feb 17, 2012 at 16:48 | comment | added | pichenettes | 1/ YIN is a classic pitch detection algorithm for speech and music signals. recherche.ircam.fr/equipes/pcm/cheveign/pss/2002_JASA_YIN.pdf . DMF is the "difference magnitude function", the quantity computed by algorithms like YIN for pitch estimation. 2/ A pitch detection algorithm like YIN will yield the estimate of the fundamental frequency, and a "confidence score" indicating how likely it is that the returned pitch is the correct answer. On noisy signals or signals exhibiting several pitches, this will be low, on a pure sine wave this will be very high. | |
| Feb 17, 2012 at 16:18 | comment | added | Spacey | @pichenettes 6) Finally, I should have mentioned that this is to be used as an 'alarm', so that if either signal-1 or signal-2 are present, I get an alarm to ring, but then nothing should go off if there is nothing there - Shouldnt there be some threshold to match before it even starts to try and classify so that you do not get false positives over nothing? (just background noise for example). (I am just learning about the Naive Bayes Classifier now, so dont know if its multi-class). 7) THANKS A LOT BY THE WAY AND THANKS IN ADVANCE! A THOUSAND AND ONE UPVOTES FOR YOU! :-) | |
| Feb 17, 2012 at 16:10 | comment | added | Spacey | @pichenettes (contd) 4) Regarding the features those exist PER WINDOW only yes? So you are computing three stds per window, of, (in this case) 101 frames? In this case when it comes time to train, my 3-D 'point' was made from 3 stds over 101 FRAMES, correct? 5) In your step 4, when you have the number 1,1,1,2,2 etc, each number corresponds to how you classified THAT window correct? The first '1' was classification of window made up of frames -50 to 50, and the second '1' from a window made up of frames -49 to 51, correct? (Window is sliding by 1 frame everytime)... | |
| Feb 17, 2012 at 15:58 | comment | added | Spacey | @pichenettes Thank you pichenettes, that is a very good answer - I have a couple follow ups though: 1) What is 'YIN' that you mention, and what is 'DMF'? I could not find anything on them through google. 2) What exactly is 'pitch confidence' as a measure? 3) You mention that you can use the auto-correlation to find pitch-confidence - autocorrelation of what, the time domain frame or the frame's STFT? (I dont understand this probably because I dont know what you mean pitch confidence). (contd...) | |
| Feb 16, 2012 at 20:56 | comment | added | pichenettes | Note that you could also do it speech recognition style if you want to leverage HTK or the like... Acoustic model: 4 phones P1, P2, P3, P4 (tone at f1, tone at f2, tone at f3, tone at f4) + 1 symbol S for silence. 1 or 2 gaussians per phone. Word model W1 for signal 1: (P1 S P2 S P3 S P4 S)+. Word model W2 for signal 2: (P1 S)+. Word model W3 for signal 3: (P1)+. Sentence model : (W1|W2|W3)*. The only caveat is that if you use a speech recognition toolbox, you'll have to tweak its feature extraction front-end since MFCCs are too low-resolution and pitch-agnostic to tell apart f1/f2/f3/f4. | |
| Feb 16, 2012 at 20:46 | history | edited | pichenettes | CC BY-SA 3.0 |
added 677 characters in body
|
| Feb 16, 2012 at 20:41 | history | answered | pichenettes | CC BY-SA 3.0 |