방법 비교
선택한 방법을 나란히 검토하세요. 서로 다른 행은 강조 표시됩니다.
| 음악 분할× | 멜로디 추출× | |
|---|---|---|
| 분야 | 음악 정보 검색 | 음악 정보 검색 |
| 계열 | Machine learning | Machine learning |
| 기원 연도≠ | 2001 | 2008 |
| 창시자≠ | Masataka Goto | Anssi Klapuri |
| 유형≠ | Audio structural analysis | Polyphonic audio analysis |
| 원전≠ | Goto, M., & Hasegawa, Y. (2001). Automatic transcription of popular music audio. In Proceedings of the Fourth International Conference on Music Information Retrieval. link ↗ | Salamon, J., & Gómez, E. (2014). Melody extraction from polyphonic music signals using pitch contour characteristics. IEEE Transactions on Audio, Speech, and Language Processing, 20(6), 1759-1770. link ↗ |
| 별칭 | structural segmentation, music structure analysis, section boundary detection | pitch contour extraction, melodic line extraction, f0 tracking |
| 관련 | 5 | 5 |
| 요약≠ | Music segmentation is the task of dividing a musical recording into distinct structural sections (e.g., verse, chorus, bridge, pre-chorus, outro). Introduced by Goto (2001), it identifies major structural boundaries and labels sections according to musical form. Segmentation is essential for music understanding, audio editing, and composition analysis. It enables higher-level tasks like cover song identification and song structure-aware music generation. | Melody extraction is the task of automatically isolating the main melodic contour from polyphonic music recordings. It originated from music transcription research in the 2000s and addresses the core challenge of human pitch perception: identifying the perceptually dominant pitch when many instruments play simultaneously. Modern approaches use deep learning and are essential for music analysis, cover song detection, and music-to-lyrics alignment. |
| ScholarGate데이터셋 ↗ |
|
|