Voice recognition client system for processing online voice recognition, voice recognition server system, and voice recognition method

A speech recognition server system for generating a speech recognition result from a sound signal, the sound signal being input into a speech recognition client system, the speech recognition server system comprising:

• a receiving unit configured to receive the sound signal, from the speech recognition client system, as a plurality of sound signal units of a predetermined length, each sound signal unit of the plurality of sound signal units being input, from the speech recognition client system, at intervals of a first predetermined unit time starting from a starting time and ending at an ending time;
• a speech recognition result generating unit configured to generate at least one intermediate result, using a partial sound signal generated by accumulating at least one of the plurality of sound signal units received by the receiving unit within a second predetermined unit time;
• and
• a transmitting unit configured to transmit the at least one intermediate result to the speech recognition client system at intervals of the second predetermined unit time,
• wherein the at least one intermediate result is displayed at a display time corresponding to a time between the starting time and the ending time.

A computer-implemented speech recognition method, the speech recognition method comprising:

• transmitting, to a speech recognition server system, a sound signal as a plurality of sound signal units of predetermined length, each sound signal unit of the plurality of sound signal units being transmitted at intervals of a first predetermined unit time starting from a starting time and ending at an ending time, and receiving at least one intermediate result from the speech recognition server system at intervals of a second predetermined unit time;
• and
• displaying, with a computer, the received at least one intermediate result at a display time, corresponding to a time between the starting time and the ending time;
• wherein the at least one intermediate result is generated, by the speech recognition server system, by accumulating at least one of the plurality of sound signal units received within the second predetermined unit time.

A computer-implemented speech recognition method of generating a speech recognition result from a sound signal, the sound signal being received from a speech recognition client system, the speech recognition method comprising:

• receiving, with a computer, the sound signal as a plurality of sound signal units of a predetermined length, each sound signal unit of the plurality of sound signal units being received, from the speech recognition client system, at intervals of a first predetermined unit time starting from a starting time and ending at an ending time;
• generating, with the computer, at least one intermediate result using a partial sound signal generated by accumulating at least one of the plurality of sound signal units received within a second predetermined unit time;
• and
• transmitting the at least one intermediate result to the speech recognition client system at intervals of the second predetermined unit time,
• wherein the at least one intermediate result is displayed at a display time corresponding to a time between a starting time and an ending time of the sound signal.

A non-transitory computer-readable medium storing a program for generating a speech recognition result from a sound signal received from a speech recognition client system, said program, when executed by a computer, causing the computer to perform steps comprising:

• receiving the sound signal as a plurality of sound signal units of a predetermined length, each sound signal unit of the plurality of sound signal units being received, from the speech recognition client system, at intervals of a first predetermined unit time starting from a starting time and ending at an ending time;
• generating at least one intermediate result using a partial sound signal generated by accumulating at least one of the plurality of sound signal units received within a second predetermined unit time;
• and
• transmitting the at least one intermediate result to the speech recognition client system at intervals of the second predetermined unit time,
• wherein the at least one intermediate result is displayed at a display time corresponding to a time between a starting time and an ending time of the sound signal.