We generalize the non-negative matrix factorization (NMF) generative model to incorporate an explicit offset. Multiplicative estimation algorithms are provided for the resulting sparse affine NMF model. We show that the affine model has improved uniqueness properties and leads to more accurate identification of mixing and sources.
|Title of host publication||ICASSP 2007, Hawaii, USA|
|Publication status||Published - 2007|
|Event||32nd IEEE International Conference on Acoustics, Speech and Signal Processing 2007 - Honolulu, HI, United States|
Duration: 15 Apr 2007 → 20 Apr 2007
|Conference||32nd IEEE International Conference on Acoustics, Speech and Signal Processing 2007|
|Period||15/04/2007 → 20/04/2007|