SOUND PROCESSING METHOD

To reduce processing load for sound source separation.SOLUTION: A sound processing system includes: an acquisition unit 231 that acquires input data D(m) including first input data representing a component of a first frequency band in a first sound corresponding to a first sound source, second input data representing a component of the first frequency band in a second sound corresponding to a second sound source different from the first sound source, and mixed sound data representing a sound including a component of a frequency band including a second frequency band different from the first frequency band in a mixed sound of the first sound and the second sound; and a generation unit 232 that by inputting the input data D(m) to an estimation model M subjected to learning, generates at least one of first output data representing a component of a frequency band including the second frequency band in the first sound and second output data representing a component of a frequency band including the second frequency band in the second sound.SELECTED DRAWING: Figure 2 【課題】音源分離のための処理負荷を軽減する。【解決手段】音響処理システムは、第１音源に対応する第１音のうち第１周波数帯域の成分を表す第１入力データと、第１音源とは異なる第２音源に対応する第２音のうち第１周波数帯域の成分を表す第２入力データと、第１音と第２音との混合音のうち第１周波数帯域とは異なる第２周波数帯域を含む周波数帯域の成分を含む音を表す混合音データと、を含む入力データＤ(m)を取得する取得部２３１と、学習済の推定モデルＭに入力データＤ(m)を入力することで、第１音のうち第２周波数帯域を含む周波数帯域の成分を表す第１出力データと、第２音のうち第２周波数帯域を含む周波数帯域の成分を表す第２出力データとの少なくとも一方を生成する生成部２３２とを具備する。【選択図】図２