Electronic device and method for recognizing speech

An electronic device, comprising:

• a memory;
• and
• a processor is configured to:
• based on receiving a plurality of audio signals generated from a sound source through a plurality of microphones, obtain a power ratio based on a power value for each of the plurality of audio signals,
• obtain first and second direction information for the sound source corresponding to first and second audio signals from among the plurality of audio signals
• identify a speech section based on difference information between the first and the second direction information and the obtained power ratio;
• process an audio signal among the audio signals included in the identified speech section;
• perform a speech recognition on the processed audio signal included in the identified speech section;
• control the electronic device to perform a function corresponding to a result of the speech section;
• and
• based on the difference between the first and the second direction information being less than a threshold, identify a starting point and an ending point of the speech section from the first audio signal and the second audio signal, respectively, and
• wherein the processing the audio signal processes the audio signal using at least one of a beam forming scheme, a geometric source separation (GSS) scheme or a blind source extraction (BSE) scheme.

A non-transitory computer-readable medium storing computer instructions, which when executed by a processor of an apparatus, causes the apparatus to perform operations comprising:

• receiving a plurality of audio signals generated from a sound source through a plurality of microphones;
• obtaining a power ratio based on a power value for each of the plurality of audio signals;
• obtaining first and second direction information for the sound source corresponding to first and second audio signals from among the plurality of audio signals;
• identifying a speech section based on difference information between the first and the second direction information and the obtained power ratio;
• process an audio signal among the audio signals included in the identified speech section;
• performing a speech recognition on the processed audio signal included in the identified speech section;
• and
• controlling the apparatus to perform a function corresponding to a result of the speech section,
• wherein the identifying the speech section identifies, based on the difference between the first and the second direction information being less than a threshold, a starting point and an ending point of the speech section from the first audio signal and the second audio signal, respectively, and
• wherein the processing the audio signal processes the audio signal using at least one of a beam forming scheme, a geometric source separation (GSS) scheme or a blind source extraction (BSE) scheme.