Matches in SemOpenAlex for { <https://semopenalex.org/work/W373715904> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W373715904 abstract "A total of 4 algorithms have been submitted. They consist in a set of Matlab functions and a binary file compiled under the Windows operating system. In accordance with the functional framework for rhythm description systems proposed in [2], the algorithms consist in three main processing blocks: feature list creation from audio, periodicity function computation and parsing. Low-level features are computed on a frame-by-frame basis. Algorithms account for 4 different sets of features. Algorithm0 uses 13 features: the magnitudenormalized derivative of the energy in the 8 frequency bands proposed by Dixon et al. [1] (note that Dixon et al. [1] use the derivative, normalizing by the magnitude yields significant improvements) and the magnitudenormalized derivative of 5 spectral features (the mean of the spectral peaks and the spectrum geometric mean, kurtosis, low-frequency energy ratio, mean and skewness). Algorithm1 uses 8 features: Algorithm0 energy features. Algorithm2 uses 9 spectral features: the magnitude-normalized derivative of the the spectral peaks mean, harmonic centroid and harmonic deviation and the spectrum flatness, geometric mean, kurtosis, low-frequency energy ratio, mean and skewness. Algorithm3 uses 13 features: the derivative of the MFCCs. For more details on the relevance of different feature sets, see [3]. All algorithms implement the autocorrelation as periodicity function. Parsing the periodicity function and inferring the most salient tempo (T1, and the meter as a byproduct) is done similarly as in [1]: prominent peaks are collected from each periodicity function. The algorithm then considers all pairs of peaks as possible beat/measure combinations, and computes the fit of all periodicity peaks to each hypothesis, using a weighted sum, where the weights represent the likelihood of each metrical unit appearing as a strong periodicity, given the meter [1]. The second most salient tempo (T2) is chosen as the periodicity (differing from T1) with the highest weight. The normalised relative salience/strength of T1 (ST1) is computed from T1 and T2 weights. The phases of T1 and T2 (P1 and P2) are computed by correlation of pulse trains with feature lists." @default.
- W373715904 created "2016-06-24" @default.
- W373715904 creator A5015454019 @default.
- W373715904 creator A5077484271 @default.
- W373715904 date "2005-01-01" @default.
- W373715904 modified "2023-09-25" @default.
- W373715904 title "Influence of input features in perceptual tempo induction" @default.
- W373715904 cites W156361483 @default.
- W373715904 cites W2029446232 @default.
- W373715904 cites W2061457313 @default.
- W373715904 cites W2181079418 @default.
- W373715904 hasPublicationYear "2005" @default.
- W373715904 type Work @default.
- W373715904 sameAs 373715904 @default.
- W373715904 citedByCount "2" @default.
- W373715904 crossrefType "journal-article" @default.
- W373715904 hasAuthorship W373715904A5015454019 @default.
- W373715904 hasAuthorship W373715904A5077484271 @default.
- W373715904 hasConcept C105795698 @default.
- W373715904 hasConcept C11413529 @default.
- W373715904 hasConcept C122342681 @default.
- W373715904 hasConcept C134306372 @default.
- W373715904 hasConcept C146599234 @default.
- W373715904 hasConcept C153180895 @default.
- W373715904 hasConcept C154945302 @default.
- W373715904 hasConcept C166963901 @default.
- W373715904 hasConcept C33257320 @default.
- W373715904 hasConcept C33923547 @default.
- W373715904 hasConcept C41008148 @default.
- W373715904 hasConcept C5297727 @default.
- W373715904 hasConceptScore W373715904C105795698 @default.
- W373715904 hasConceptScore W373715904C11413529 @default.
- W373715904 hasConceptScore W373715904C122342681 @default.
- W373715904 hasConceptScore W373715904C134306372 @default.
- W373715904 hasConceptScore W373715904C146599234 @default.
- W373715904 hasConceptScore W373715904C153180895 @default.
- W373715904 hasConceptScore W373715904C154945302 @default.
- W373715904 hasConceptScore W373715904C166963901 @default.
- W373715904 hasConceptScore W373715904C33257320 @default.
- W373715904 hasConceptScore W373715904C33923547 @default.
- W373715904 hasConceptScore W373715904C41008148 @default.
- W373715904 hasConceptScore W373715904C5297727 @default.
- W373715904 hasLocation W3737159041 @default.
- W373715904 hasOpenAccess W373715904 @default.
- W373715904 hasPrimaryLocation W3737159041 @default.
- W373715904 hasRelatedWork W1797318752 @default.
- W373715904 hasRelatedWork W1991656160 @default.
- W373715904 hasRelatedWork W2024600556 @default.
- W373715904 hasRelatedWork W2036266388 @default.
- W373715904 hasRelatedWork W2040292825 @default.
- W373715904 hasRelatedWork W2066156173 @default.
- W373715904 hasRelatedWork W2077805670 @default.
- W373715904 hasRelatedWork W2085370744 @default.
- W373715904 hasRelatedWork W2114002670 @default.
- W373715904 hasRelatedWork W2114907815 @default.
- W373715904 hasRelatedWork W2135444961 @default.
- W373715904 hasRelatedWork W2160225526 @default.
- W373715904 hasRelatedWork W2167364061 @default.
- W373715904 hasRelatedWork W2181998524 @default.
- W373715904 hasRelatedWork W2230675850 @default.
- W373715904 hasRelatedWork W2583655541 @default.
- W373715904 hasRelatedWork W298430120 @default.
- W373715904 hasRelatedWork W3100396845 @default.
- W373715904 hasRelatedWork W39263461 @default.
- W373715904 hasRelatedWork W1487762982 @default.
- W373715904 isParatext "false" @default.
- W373715904 isRetracted "false" @default.
- W373715904 magId "373715904" @default.
- W373715904 workType "article" @default.