| 摘要 |
An audio generation method comprises: acquiring comparison audio, performing theme extraction on the comparison audio to obtain a comparison note sequence comprising comparison note positions, comparison note pitches, and comparison note durations; acquiring original audio matching the comparison audio via audio retrieval, and acquiring an original note sequence corresponding to the original audio, wherein the original note sequence is obtained via performing theme extraction on the original audio, and comprises original note positions, original note pitches, and original note durations; calculating, according to the comparison note sequence and the original note sequence, theme distances between segments of the comparison audio and segments of the original audio; and taking a segment of the original audio having the smallest theme distance as generated audio. |
