Please use this identifier to cite or link to this item: https://scholarbank.nus.edu.sg/handle/10635/233974
Title: AUTOMATIC LYRICS TRANSCRIPTION OF POLYPHONIC MUSIC
Authors: GAO XIAOXUE
Keywords: lyrics transcription, music information processing, music information retrieval
Issue Date: 9-May-2022
Citation: GAO XIAOXUE (2022-05-09). AUTOMATIC LYRICS TRANSCRIPTION OF POLYPHONIC MUSIC. ScholarBank@NUS Repository.
Abstract: Automatic Lyrics Transcription of polyphonic music (ALTP) aims to recognize the lyrics from singing vocals in the presence of music accompaniment. Despite considerable research effort, there exist several challenges in ALTP, such as complicated multi-step training, imperfect singing vocal extraction, complex structure of polyphonic music, music genre discrepancy and background music interference problems. This thesis advances deep learning techniques and E2E frameworks to overcome the challenges. The first contribution is to remedy the imperfect singing vocal extraction by the proposed music-robust model. The second contribution focuses on overcoming the multi-step training problem by firstly building E2E frameworks using transformer. We further address the complex polyphonic music problem by formulating multi-transcribers with chords to interpret music. The third contribution resolves the music genre discrepancy problem by proposing a genre-specific adapter to handle various singing styles across genres. The final contribution attempts to address background music interference problem via E2E joint-training approaches.
URI: https://scholarbank.nus.edu.sg/handle/10635/233974
Appears in Collections:Ph.D Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
Gao_Xiaoxue.pdf1.75 MBAdobe PDF

OPEN

NoneView/Download

Page view(s)

32
checked on Feb 2, 2023

Download(s)

6
checked on Feb 2, 2023

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.