SNR of signal
in [Matlab]
|
Prev: Elliptic PDEtoolbox trouble!!
Next: Results not repeatable for sum() function on fixed / deterministic vector
From: Steven Lee on 5 Apr 2010 14:59 Hi guys. I'm currently working on a sound project which requires me to reduce noise from noisy speech signal. I'm using wav files taken from the NOISEUS database which provides the clean speech and the noise currupted signal (Clean speech + Noise) in different SNR of 15 10 5 and 0 dB. And after reducing the noise from the noisy signal I want to know the SNR of the output i've generated. Is there anyways to find out the new SNR with only the clean and noisy speech signal(this is for the testing part I want to know if the algorithm used actually improve the SNR of the noisy speech signal)? I dont have any clue about the noise they've merged the clean signal with. If someone know the solution please let me know. Thanks! (^.^)b
From: Greg Heath on 5 Apr 2010 22:45
On Apr 5, 2:59 pm, "Steven Lee" <slayerti...(a)aol.com> wrote: > Hi guys. > > I'm currently working on a sound project which requires me to reduce noise from noisy speech signal. > > I'm using wav files taken from the NOISEUS database which provides the clean speech and the noise currupted signal (Clean speech + Noise) in different SNR of 15 10 5 and 0 dB. > > And after reducing the noise from the noisy signal I want to know the SNR of the output i've generated. Is there anyways to find out the new SNR with only the clean and noisy speech signal(this is for the testing part I want to know if the algorithm used actually improve the SNR of the noisy speech signal)? I dont have any clue about the noise they've merged the clean signal with. > > If someone know the solution please let me know. 1. Subtract the given clean signal, S0, from the given noisy signal S = S0 +N0 to estimate N0 2. Calculate SNR0 and compare with the label 3. Reduce the noise to obtain the new signal S1 = S0 + N1 4. Subtract S0 to estimate N1 5. Estimate SNR1 and compare with SNR0 Hope this helps. Greg |