Online target-speech extraction method based on auxiliary function for robust automatic speech recognition

A target speech signal extraction method of extracting a target speech signal from input signals input to at least two or more microphones for robust speech recognition, by a processor of a speech recognition apparatus, comprising:

• (a) initializing a steering vector for a target speech source and an adaptive vector (w1(k)) for the target speech source, setting a real output channel of the target speech source as an output by the adaptive vector for the target speech source, initializing adaptive vectors (w2(k), . . . , wM(k)) for a noise and setting a dummy channel as an output by the adaptive vectors for the noise;
• (b) setting a cost function for minimizing dependency between a real output for the target speech source and a dummy output for the noise using independent component analysis (ICA) or independent vector analysis (IVA);
• (c) setting an auxiliary function to the cost function, and updating the adaptive vector (w1(k)) for the target speech source and the adaptive vectors (w2(k), . . . , wM(k)) for the noise by using the auxiliary function and the steering vector for the target speech source;
• (d) estimating the target speech signal by using the adaptive vector for the target speech source thereby extracting the target speech signal from the input signals;
• and
• (e) updating the steering vector for the target speech source;
• wherein the (b)˜(e) are performed repeatedly;
• and
• wherein the auxiliary function is set an inequality relation so that the auxiliary function has always values greater than or same as that of the cost function.