On the approximate W-disjoint orthogonality of speech

By Scott T. Rickard, Özgür Yilmaz. IEEE International Conference on Acoustics, Speech, and Signal Processing, 2002

312 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Multi-Speaker DOA Estimation Using Deep Convolutional Networks Trained With Noise Signals (2018, methodology): “To overcome this problem, the property of W-disjoint orthogonality [26], which holds approximately for speech signals, was utilized.”
Multi-Channel Deep Clustering: Discriminative Spectral and Spatial Embeddings for Speaker-Independent Speech Separation (2018, methodology): “The DUET algorithm [6, 7] assumes that the microphone pairs are placed suf-ﬁciently close to each other so that phase-wrapping effects can be neglected.”
Blind separation of speech mixtures via time-frequency masking (2004, background): “Empirical support for this can be found in [7] and [13], and a more extensive discussion is given in Section II-A.”
A time-frequency blind signal separation method applicable to underdetermined mixtures of dependent sources (2005, background): “only one source should occur in each TF window) which is quite restrictive, even if it has been shown that approximate W-disjoint orthogonality is sufficient for most speech signal [16,17].”
A Survey of Sound Source Localization with Deep Learning Methods (2021, methodology): “The model parameters are estimated with histogram-based or Expectation-Maximization algorithms exploiting the sparsity of sound sources in the time-frequency domain [90].”