rong gong profile photo

Rong Gong

Working from home
Vienna
Austria

woshicoolbe<at>gmail.com

CV in English

contact details

Linkedin Google Scholar Blogger Github

From September 2019, I work as a research scientist in Nuance Communications, Vienna, developing multichannel array far-field speech recognition technologies.

PhD thesis

Companion webpage, Automatic assessment of singing voice pronunciation: a case study with jingju music

Research interests

  • Music information retrieval MIR
  • Audio, music and speech signal processing and machine learning
  • Sound scene classification and sound event detection
  • Psychoacoustics

Academic activities

  • Google Speech Summit traveling grant, London, May 2018
  • Task captain of MIREX 2018: Automatic Lyrics-to-Audio Alignment (Link)
  • ISMIR conference reviewer 2017, 2018, 2019
  • DCASE workshop reviewer 2019

Education

  • PhD Cum laude, Music Technology Group - Universitat Pompeu Fabra MTG-UPF, Barcelona
  • Master in Acoustics, Signal Processing, and Computer Science Applied to Music (ATIAM), UPMC/IRCAM/Telecom Paristech, Paris
  • Master in Acoustics, Psychoacoustics and Signal Processing, Communication university of China, Beijing
  • B.Math in Applied Mathematics from Northwest university, Xi'an, China

Projects

  • Essentia, github page Implement PYin (probabilistic Yin) pitch extraction algorithms in Essentia (Open-source library and tools for audio and music analysis, description and synthesis).
  • TECSOME-MusicCritic, demo page Develop automatic solfè€ge assessment API using KALDI.
  • Remorse at death iOS app, github page Project collaborator-Beijing City Social Science Foundation 16JDYTA016; Collect jingju singing recording materials of both professional and amateur singers; Develop iOS jingju singing courseware.
  • CompMusic, "Computational Models for the Discovery of the World's Music", funded by the European Research Council (2011~2017).
    • Jingju (Beijing opera) singing voice a cappella dataset, part1, part2, part3
    • Jingju syllable and phoneme onset detection and segmentation, baseline models, improved models
    • Jingju mispronunciation detection, code
    • Jingju pronunciation and overall quality similarity, code
  • Antescofo Improve the performance of the score following system - Antescofo on singing voice.

Publications

list of publications