English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

We present Vibrato Nonnegative Tensor Factorization, an algorithm forsingle-channel unsupervised audio source separation with an application toseparating instrumental or vocal sources with nonstationary pitch from musicrecordings. Our approach extends Nonnegative Matrix Factorization for audiomodeling by including local estimates of frequency modulation as cues in theseparation. This permits the modeling and unsupervised separation of vibrato orglissando musical sources, which is not possible with the basic matrixfactorization formulation.  The algorithm factorizes a sparse nonnegative tensor comprising the audiospectrogram and local frequency-slope-to-frequency ratios, which are estimatedat each time-frequency bin using the Distributed Derivative Method. The use oflocal frequency modulations as separation cues is motivated by the principle ofcommon fate partial grouping from Auditory Scene Analysis, which hypothesizesthat each latent source in a mixture is characterized perceptually by coherentfrequency and amplitude modulations shared by its component partials. We derivemultiplicative factor updates by Minorization-Maximization, which guaranteesconvergence to a local optimum by iteration. We then compare our method to thebaseline on two separation tasks: one considers synthetic vibrato notes, whilethe other considers vibrato string instrument recordings.

Nonnegative tensor factorization with frequency modulation cues for blind audio source separation