Automated difference recognition between speaking sounds and music
View Patent ↗The present invention relates to means and methods of automated difference recognition between speech and music signals in voice communication systems, devices, telephones, and methods, and more specifically, to systems, devices, and methods that automate control when either speech or music is detected over communication links. The present invention provides a novel system and method for monitoring the audio signal, analyze selected audio signal components, compare the results of analysis with a pre-determined threshold value, and classify the audio signal either as speech or music.
1. A method of manipulating sound signals, the method comprising the steps of:
a) obtaining ‘N_DEC’ number of buffers, each buffer having N number of samples of sound signals;
b) passing each buffer of N samples through a high pass filter (HPF) and obtaining output signals;
c) finding power of HPF output signals for each of the N samples;
d) averaging the power over a period of time to obtain a power level;
e) using the signals being passed through the HPF in a Goertzel algorithm to compute a sequence s(n);
f) using the sequence s(n) to compute digital Fourier transforms (DFTs) at different frequencies (ω);
g) storing the DFTs in an array of N_DEC×M, wherein N_DEC is number of buffers and M is number of preselected frequencies of sound signals, the sound signals comprising musical notes;
h) calculating histograms for each preselected frequencies of sound signals;
i) summing up histogram bins based on the histograms with a higher value as compared to a preselected threshold and storing them in a 1×M Cases Array;
j) calculating the difference signal by taking the first difference between adjacent elements in the cases array;
k) calculating a standard deviation of the difference signal;
l) selecting a bottom threshold for the power level;
m) declaring the signal as music signal if the deviation is above the bottom threshold; and
n) declaring the signal as speech or pause signal if the deviation is below the bottom threshold.
2. The method of claim 1 wherein N is between 128 to 8192 samples.
3. The method of claim 2 wherein N_DEC is between 5 to 950 buffers.
4. The method of claim 3 wherein k, a time index, is between the values of 1 and N and is defined within an equation used for finding the power of the buffer of N samples as:
pwr
=
1
N
∑
k
=
0
N
x
(
k
)
*
x
(
k
)
.
5. The method of claim 4 wherein M, the number of pre-selected frequencies of musical notes is in the range of 1 to 900.
6. The method of claim 5 wherein the pre-selected frequencies of musical notes are in the frequency ranges of 2 Hz to 40,000 Hz.