METHOD AND APPRATUS FOR TRAINING ACOUSTIC SCENE RECOGNITION MODEL AND METHOD AND APPRATUS FOR RECONITION OF ACOUSTIC SCENE USING ACOUSTIC SCENE RECOGNITION MODEL

摘要

Disclosed are a method and apparatus for learning an acoustic scene recognition model, and a method and apparatus for recognizing an acoustic scene through an acoustic scene recognition model. A method for learning an acoustic scene recognition model for recognizing an acoustic scene based on an acoustic event according to an embodiment of the present invention is to obtain characteristic information of the acoustic signal from the acoustic signal tagged with a label related to the acoustic event and a label related to the acoustic scene. extracting over time; training an acoustic event recognition model using the extracted feature information and a label related to the acoustic event; obtaining acoustic event information corresponding to the characteristic information by inputting the characteristic information into the trained acoustic event recognition model; combining adjacent feature information in chronological order among the feature information, and merging the acquired sound event information into the combined feature information; and training an acoustic scene recognition model using the feature information in which the acoustic event information is merged and a label related to the acoustic scene.