Wavelets and subband coding
Learning Overcomplete Representations
Neural Computation
Blind separation of disjoint orthogonal signals: demixing N sources from 2 mixtures
ICASSP '00 Proceedings of the Acoustics, Speech, and Signal Processing, 2000. on IEEE International Conference - Volume 05
| Hi-index | 0.00 |
Source sparsity is a common assumption in many solutions proposed in literature to the problem of blind source separation with more sources than mixtures. As shown in this work, representation of signals in different wavelet domains can be efficiently applied in order to get improved sparsity. Moreover, the approach here presented allows to directly perform a de-noising operation after the separation algorithm, at a very low computational cost, resulting in a further improvement of source recovering when noise is present at mixture level. Experimental results confirm the effectiveness of developed idea.