Companion webpage for the PhD thesis of Rong Gong

Thesis Zenodo.org

Corpora and datasets

Access to the corpora and datasets will be through the Zenodo.org and MusicBrainz.org

Research corpus

  1. Jingju a cappella singing voice dataset part 1
  2. Jingju a cappella singing voice dataset part 2
  3. Jingju a cappella singing voice dataset part 3
  4. Jingju a cappella singing voice dataset metadata on MusicBrainz (organized by Honglin Ma)

Test dataset

  1. Automatic syllable and phoneme segmentation test dataset part 1 - ASPS1
  2. Automatic syllable and phoneme segmentation test dataset part 2 - ASPS2
  3. Pronunciation and overall quality similarity measures test dataset - POQSM

Code

The links to code related to the thesis are listed. Up-to-date links to code (including future releases) will be available on https://github.com/ronggong

  1. Automatic syllable and phoneme segmentation baseline code
  2. Automatic syllable and phoneme segmentation onset detection function improvement code
  3. Mispronunciation detection code
  4. Pronunciation and overall quality similarity measures code