[en] Applying the word2vec technique, commonly used in language modeling, to
melodies, where notes are treated as words in sentences, enables the capture of
pitch information. This study examines two datasets: 20 children's songs and an
excerpt from a Bach sonata. The semantic space for defining the embeddings is
of very small dimension, specifically 2. Notes are predicted based on the 2, 3
or 4 preceding notes that establish the context. A multivariate analysis of the
results shows that the semantic vectors representing the notes have a multiple
correlation coefficient of approximately 0.80 with their pitches.
Disciplines :
Computer science
Author, co-author :
Defays, Daniel ; Université de Liège - ULiège > Département de Psychologie
Language :
English
Title :
From melodic note sequences to pitches using word2vec
Alternative titles :
[fr] Des mélodies aux hauteurs des notes à l’aide de word2vec
Original title :
[en] From melodic note sequences to pitches using word2vec
