| 摘要 |
The present invention provides a speaker identification method for a speaker identification system. This system stores: a plurality of registered speech signals respectively generated on the basis of speech of a plurality of speakers; and a plurality of speaker images which are respectively associated with the plurality of registered speech signals, and which respectively represent the plurality of speakers. In this method, speech of a speaker located in the vicinity of a display is acquired, a speaker-speech signal is generated from the speech of the speaker, a registered speech signal corresponding to the speaker-speech signal is identified from among the plurality of stored registered speech signals, and a speaker image associated with the identified registered speech signal is displayed on the display for a time period in which at least the speech of the speaker serving as the source from which the speaker-speech signal was generated is being acquired. |
