Person's i.e. speaker, actual speech duration and percentage speech duration detecting method for use in e.g. meeting, involves detecting and processing acoustic signal for determining frequency of break in number per time unit

摘要

The method involves detecting and processing an acoustic signal for identifying a speaker and for determining actual speech duration and a percentage speech duration as well as the frequency of the break in number per time unit from a predetermined time point as well as for delivering the determined duration and frequency to a display. A direction of an incoming signal is detected with the aid of the distributed arrangement of direction-sensitive microphones. A classification process is utilized with a training phase and an operating phase for the identification of the speaker. An independent claim is also included for a device for detecting an actual speech duration and a percentage speech duration of a person.