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.
TODO: could be done weasier with this code, and check last method in Word
grTruthWordList = testT(decoder.lyricsWithModels)