One aspect of the present disclosure provides a method for automatically removing a background sound source of audio data of a video. The method for removing a background sound source comprises the steps of: separating audio data of a video comprising at least one sound source component into a first component relating to a human voice and a second component relating to sounds other than the human voice, by means of a first separation model; separating the first component into a vocal component and a speech component by means of a second separation model; separating the second component into a music component and a noise component by means of a third separation model; and generating audio data having the background sound source removed from the audio data of the video, by synthesizing the speech component and the noise component.