Note: This bibliographic page is archived and will no longer be updated. For an up-to-date list of publications from the Music Technology Group see the Publications list .

Automatic Alignment of Long Syllables In A cappella Beijing Opera

Title Automatic Alignment of Long Syllables In A cappella Beijing Opera
Publication Type Conference Paper
Year of Publication 2016
Conference Name 6th International Workshop on Folk Music Analysis (FMA 2016)
Authors Dzhambazov, G. , Yang Y. , Caro Repetto R. , & Serra X.
Pagination 88-91
Conference Start Date 15/06/2016
Conference Location Dublin, Ireland
Abstract In this study we propose how to modify a standard approach for text-to-speech alignment to apply in the case of alignment of lyrics and singing voice. We model phoneme durations by means of a duration-explicit hidden Markov model (DHMM) phonetic recognizer based on MFCCs. The phoneme durations are empirically set in a probabilistic way, based on prior knowledge about the lyrics structure and metric principles, specific for the Beijing opera music tradition. Phoneme models are GMMs trained directly on a small corpus of annotated singing voice. The alignment is evaluated on a cappella material from Beijing opera, which is characterized by its particularly long syllable durations. Results show that the incorporation of music-specific knowledge results in a very high alignment accuracy, outperforming significantly a baseline HMM-based approach.
preprint/postprint document http://arrow.dit.ie/fema/1/
Additional material:

Dataset

The dataset consist of excerpts from 15 arias of two female singers. You can access the annotations of the dataset at http://compmusic.upf.edu/node/286 . Please refer to Section 5 in the paper for the statistics of the dataset.

Code

An efficient open-source python implementation together with documentation is available at
https://github.com/georgid/AlignmentDuration/tree/noteOnsets/src/for_jingju

Demo Video

https://vimeo.com/174680957