Sound source separation system, sound source separation method, and acoustic signal acquisition device
The invention provides a sound source separation system, a sound source separation method, and an acoustic signal acquisition device which can precisely separate a target sound and a disturbance sound coming from an arbitrary direction, and which ensures miniaturization of a device. A sound source separation system 10 comprises two microphones 21, 22 disposed side by side in a direction in which a target sound comes from, a target sound superior signal generator 30 which performs a linear combination process for emphasizing the target sound, using the received sound signals of the microphones to generate a target sound superior signal, a target sound inferior signal generator 40 which performs a linear combination process for suppressing the target sound, using the received sound signals of the microphones 21, 22 , to generate a target sound inferior signal, and a separation unit 60 which separates the target sound and a disturbance sound, using a target sound superior signal spectrum and a target sound inferior signal spectrum.
1. A sound source separation system that separates a target sound and a disturbance sound coming from an arbitrary direction other than a direction from which the target sound comes, comprising:
a plurality of different-directional-signal-group generators each generating more than or equal to two combinations of spectrums of a plurality of signals each of which has a different directivity, using received sound signals of a plurality of microphones; and
a sensitive region formation unit which determines whether or not a relationship between powers of the spectrums in a combination simultaneously satisfies a plurality of conditions each defined for a combination, for each frequency band, using more than or equal to two combinations of the spectrums of the plurality of signals generated by the respective different-directional-signal-group generators, and performs multidimensional band selection of assigning power of a spectrum selected beforehand to a spectrum of the target sound to be separated, for a frequency band where the plurality of conditions are simultaneously satisfied, wherein
each different-directional-signal-group generator generates a spectrum of a target sound superior signal and a spectrum of a target sound inferior signal using the received sound signals of the plurality of microphones,
the sensitive region formation unit sets a condition for each combination as a condition that power of the spectrum of the target sound superior signal is larger than power of the spectrum of the target sound inferior signal, and determines whether or not those conditions are simultaneously satisfied for each frequency band,
the sound source separation system has a total of three first, second and third microphones disposed at respective vertices of a triangle,
a first different-directional-signal-group generator including
a first target sound superior signal generator which acquires a difference between a received sound signal of the first microphone and a received sound signal of the second microphone undergone a delayed process on a time domain or a frequency domain and generates a first target sound superior signal,
a second target sound superior signal generator which acquires a difference between a received sound signal of the second microphone and a received sound signal of the first microphone undergone a delayed process on a time domain or a frequency domain, and generates a second target sound superior signal,
a target sound inferior signal generator which acquires a difference between received sound signals of the first and second microphones on a time domain or a frequency domain, and
an integration unit which compares powers for each frequency band using a spectrum of the first target sound superior signal generated by the first target sound superior signal generator or obtained by a subsequent frequency analysis and a spectrum of the second target sound superior signal generated by the second target sound superior signal generator or obtained by a subsequent frequency analysis, and performs a spectrum integration process of assigning inferior power to a spectrum of a target sound superior signal,
a second different-directional-signal-group generator including
a first target sound superior signal generator which acquires a difference between a received sound signal of the third microphone and a received sound signal of the second microphone undergone a delayed process on a time domain or a frequency domain and generates a first target sound superior signal,
a second target sound superior signal generator which acquires a difference between a received sound signal of the second microphone and a received sound signal of the third microphone undergone a delayed process on a time domain or a frequency domain, and generates a second target sound superior signal,
a target sound inferior signal generator which acquires a difference between received sound signals of the second and third microphones on a time domain or a frequency domain, and
an integration unit which compares powers for each frequency band using a spectrum of the first target sound superior signal generated by the first target sound superior signal generator or obtained by a subsequent frequency analysis and a spectrum of the second target sound superior signal generated by the second target sound superior signal generator or obtained by a subsequent frequency analysis, and performs a spectrum integration process of assigning inferior power to a spectrum of a target sound superior signal, and
the sensitive region formation unit performs two-dimensional-band selection of assigning power of a spectrum of a target sound superior signal generated by either one of the first and second different-directional-signal-group generators to a spectrum of the target sound to be separated.