Lyrics-to-audio-alignement system. Based on Machine Learning Algorithms: Hidden Markov Models with Viterbi forced alignment. The alignment is explicitly aware of durations of musical notes. The phonetic model are classified with MLP Deep Neural Network.
Final release after the submission of the PhD.
In the future this repository will not be supported as the author is no longer at Music Technology Group.
See the README.md if you would like to obtain a more updated version.
Duration model ready for testing