Source sound separator with spectrum analysis through linear combination and method therefor
In a source sound separator, first and second target sound predominant spectra are generated respectively by first and second processing operations for linear combination for emphasizing the target sound, using received sound signals of two microphones arrayed at a distance from each other. A target sound suppressed spectrum is generated by processing for linear combination for suppression of the target sound, using the two received sound signals. Further, a phase signal containing a larger amount of signal components of the target sound and exhibiting directivity in the direction of the target sound is generated by processing of linear combination, using the two received sound signals. The target sound and the interfering sound are separated from each other using the first and second target sound predominant spectra, the target sound suppressed spectrum, and the phase signal.
1. A source sound separator for separating a target sound and an interfering sound from each other, the interfering sound incoming from a direction different from a direction of the target sound incoming, said separator comprising:
one or more circuits configured to implement:
a first spectrum generator for using received first and second sound signals of respective first and second ones of a plurality of microphones arranged spaced apart from one another to subtract from a value for the first sound signal a value obtained by delaying the second sound signal by a first predetermined period of time in a time or frequency domain to thereby generate at least one first target sound predominant spectrum;
a second spectrum generator for subtracting from a value for the second sound signal a value obtained by delaying the first sound signal by a second predetermined period of time in the said time or frequency domain to thereby generate at least one second target sound predominant spectrum;
a third spectrum generator for using the received first and second sound signals to perform processing for linear combination for suppressing the target sound in said time or frequency domain to thereby generate at least one target sound suppressed spectrum which is to form a set with the first and second target sound predominant spectra;
a phase generator for using received sound signals of ones of the plurality of microphones to add the received sound signals in the frequency domain to thereby generate a phase signal; and
a target sound separator for using the first target sound predominant spectrum, the second target sound predominant spectrum, the target sound suppressed spectrum and the phase signal to separate the target sound and the interfering sound from each other.
2. A method for separating a target sound and an interfering sound from each other, the interfering sound incoming from a direction different from a direction of the target sound incoming, said method comprising:
preparing a first spectrum generator, a second spectrum generator, a third spectrum generator, a phase generator and a target sound separator;
generating by the first spectrum generator at least one first target sound predominant spectrum by using received first and second sound signals of respective first and second ones of a plurality of microphones arranged spaced apart from one another to subtract from a value for the first sound signal a value obtained by delaying the second sound signal by a first predetermined period of time in a time or frequency domain;
generating by the second spectrum generator at least one second target sound predominant spectrum by subtracting from a value for the second sound signal a value obtained by delaying the first sound signal by a second predetermined period of time in said time or frequency domain;
performing by the third spectrum generator processing for linear combination for suppressing the target sound in said time or frequency domain by using the received first and second sound signals to thereby generate at least one target sound suppressed spectrum which is to form a set with the first and second target sound predominant spectra;
using by the phase generator received sound signals of ones of the plurality of microphones to add the received sound signals in the frequency domain to thereby generate a phase signal; and
separating by the target sound separator the target sound and the interfering sound from each other by using the first target sound predominant spectrum, the second target sound predominant spectrum, the target sound suppressed spectrum and the phase signal.
3. A non-transitory computer-readable storage medium having a computer program recorded and separating, when installed in and executed on a computer, a target sound and an interfering sound from each other, the interfering sound incoming from a direction different from a direction of the target sound incoming, said program allowing the computer to operate as:
a first spectrum generator for using received first and second sound signals of respective first and second ones of a plurality of microphones arranged spaced apart from one another to subtract from a value for the first sound signal a value obtained by delaying the second sound signal by a first predetermined period of time in a time or frequency domain to thereby generate at least one first target sound predominant spectrum;
a second spectrum generator for subtracting from a value for the second sound signal a value obtained by delaying the first sound signal by a second predetermined period of time in said time or frequency domain to thereby generate at least one second target sound predominant spectrum;
a third spectrum generator for using the received first and second sound signals to perform processing for linear combination for suppressing the target sound in said time or frequency domain to thereby generate at least one target sound suppressed spectrum which is to form a set with the first and second target sound predominant spectra;
a phase generator for using received sound signals of ones of the plurality of microphones to add the received sound signals in the frequency domain to thereby generate a phase signal; and
a target sound separator for using the first target sound predominant spectrum, the second target sound predominant spectrum, the target sound suppressed spectrum and the phase signal to separate the target sound and the interfering sound from each other.