M. Girolami, SYMMETRICAL ADAPTIVE MAXIMUM-LIKELIHOOD-ESTIMATION FOR NOISE CANCELLATION AND SIGNAL SEPARATION, Electronics Letters, 33(17), 1997, pp. 1437-1438
The symmetric adaptive maximum likelihood estimation (SAMLE) algorithm
is proposed. It is shown to be generalisation of the symmetric adapti
ve decorrelation (SAD) algorithm for noise cancellation and signal sep
aration. Both the SAD and SAMLE algorithms are applied to the separati
on of a mixture of natural speech recorded in a realistic acoustic env
ironment. A comparative simulation confirms that the SAMLE algorithm p
rovides superior separation capabilities.