公开(公告)号：US11317200B2

公开(公告)日：2022-04-26

申请号：US17266397

申请日：2019-07-10

Applicant: UNIVERSITY OF YAMANASHI

Inventor： Kenji Ozawa

IPC: H04R1/40 ,  G10L21/0272 ,  G10L25/51

Abstract: A sound source separation system includes: a controller that: acquires pieces of sound collection data with microphones that collect sounds output from first and second sound sources. The first sound source is at a first position at which effective distances from the microphones are equal and the second sound source is at a different position. The controller further acquires, based on the sound collection data, frequency spectra in two dimensions of a circumferential direction of a circle and a time direction. The first position is a center of the circle and each of the effective distances is a radius of the circle. The controller further separates, from the frequency spectra, a first sound source spectrum and a second sound source spectrum.

公开(公告)号：US20210297773A1

公开(公告)日：2021-09-23

申请号：US17266397

申请日：2019-07-10

Applicant: UNIVERSITY OF YAMANASHI

Inventor： Kenji Ozawa

IPC: H04R1/40 ,  G10L25/51 ,  G10L21/0272

Abstract: A sound source separation system includes: a controller that: acquires pieces of sound collection data with microphones that collect sounds output from first and second sound sources. The first sound source is at a first position at which effective distances from the microphones are equal and the second sound source is at a different position. The controller further acquires, based on the sound collection data, frequency spectra in two dimensions of a circumferential direction of a circle and a time direction. The first position is a center of the circle and each of the effective distances is a radius of the circle. The controller further separates, from the frequency spectra, a first sound source spectrum and a second sound source spectrum.