Title | ||
---|---|---|
Harmonic-Temporal Clustering via Deterministic Annealing EM Algorithm for Audio Feature Extraction |
Abstract | ||
---|---|---|
ABSTRACT This paper proposes “harmonic-temporal structured clustering (HTC) method”, that allows simultaneous estimation of pitch, intensity, onset, duration, etc., of each underlying source in multi-stream audio signal, which we expect to be an effective feature extraction for MIR systems. STC decomposes,the energy patterns diffused in timefrequency space, i.e., a time series of power spectrum, into distinct clusters such that each of them is originated from a single sound stream. It becomes,clear that the problem is equivalent to geometrically approximating,the observed time series of power spectrum by superimposed,harmonictemporal structured models (HTMs), whose parameters are directly associated with the specific acoustic characteristics. The update equations in DA(Deterministic Annealing)EM algorithm for the optimal parameter convergence are derived by formulating the model with Gaussian kernel representation. The experiment showed,promising results, and verified the potential of the proposed method. Keywords: audio feature extraction, multi-pitch estima- |
Year | Venue | Keywords |
---|---|---|
2005 | International Symposium/Conference on Music Information Retrieval | gaussian kernel,power spectrum,time series,em algorithm,feature extraction |
Field | DocType | Citations |
Pattern recognition,Computer science,Expectation–maximization algorithm,Harmonic,Speech recognition,Feature extraction,Deterministic annealing,Artificial intelligence,Cluster analysis,Machine learning | Conference | 7 |
PageRank | References | Authors |
0.76 | 4 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hirokazu Kameoka | 1 | 801 | 79.06 |
Takuya Nishimoto | 2 | 227 | 28.95 |
Shigeki Sagayama | 3 | 1217 | 137.97 |