DocumentCode :
2706528
Title :
Identifying `Cover Songs´ with Chroma Features and Dynamic Programming Beat Tracking
Author :
Ellis, Daniel P. W. ; Poliner, G.E.
Author_Institution :
Dept. of Electr. Eng., Columbia Univ., New York, NY, USA
Volume :
4
fYear :
2007
fDate :
15-20 April 2007
Abstract :
Large music collections, ranging from thousands to millions of tracks, are unsuited to manual searching, motivating the development of automatic search methods. When different musicians perform the same underlying song or piece, these are known as `cover´ versions. We describe a system that attempts to identify such a relationship between music audio recordings. To overcome variability in tempo, we use beat tracking to describe each piece with one feature vector per beat. To deal with variation in instrumentation, we use 12-dimensional `chroma´ feature vectors that collect spectral energy supporting each semitone of the octave. To compare two recordings, we simply cross-correlate the entire beat-by-chroma representation for two tracks and look for sharp peaks indicating good local alignment between the pieces. Evaluation on several databases indicate good performance, including best performance on an independent international evaluation, where the system achieved a mean reciprocal ranking of 0.49 for true cover versions among top-10 returns.
Keywords :
audio signal processing; dynamic programming; music; automatic search methods; beat-by-chroma representation; chroma feature vectors; chroma features; cover songs; dynamic programming beat tracking; music audio recordings; music collections; spectral energy; Audio recording; Digital audio players; Dynamic programming; Instruments; Multiple signal classification; Music; Robustness; Search methods; Signal analysis; Spatial databases; Acoustic signal analysis; Correlation; Database searching; Dynamic programming; Music;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on
Conference_Location :
Honolulu, HI
ISSN :
1520-6149
Print_ISBN :
1-4244-0727-3
Type :
conf
DOI :
10.1109/ICASSP.2007.367348
Filename :
4218379
Link To Document :
بازگشت