Abstract | ||
---|---|---|
In order to automatically extract the main melody contours from polyphonic music especially vocal melody songs, we present an effective approach based on a Bayesian framework. According to various information from the music signals, we use a pitch evolution model describing how pitch contour changes and an acoustic model representing the acoustic characteristics when the pitch is a hypothesized one, and obtain the optimal melody contour utilizing a Viterbi algorithm. The experimental results on the RWC popular music database indicate that the overall accuracy achieves 73.4%. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1109/IIH-MSP.2014.147 | IIH-MSP |
Keywords | Field | DocType |
vocal polyphonic music,vocal melody songs,optimal melody contour,pitch contour,acoustic model,bayesian framework,music, viterbi algorithm,melody extraction,information retrieval,automatic main melody contour extraction,pitch evolution model, polyphonic music,rwc popular music database,acoustic signal processing,feature extraction,bayes methods,viterbi algorithm,audio databases, bayesian framework,polyphonic music,music signals | Pitch contour,Vocal Melody,Pattern recognition,Computer science,Speech recognition,Popular music,Artificial intelligence,Polyphony,Viterbi algorithm,Bayesian probability,Acoustic model | Conference |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Liming Song | 1 | 14 | 3.08 |
Ming Li | 2 | 4 | 0.79 |
Yonghong Yan | 3 | 656 | 114.13 |