摘要 |
The system has a processor (14) for decomposing an audio signal received on a sensor (10) into representative vectors of a speaker. The processor authenticates and transforms a super-vector, which is derived from concatenation of the representative vectors, into binary data. The processor transmits the obtained binary data to an extraction module (51) for extracting authentication of the speaker, so as to authenticate the speaker and/or to generate a given cryptographic key associated with the speaker. An independent claim is also included for a method for audio authentication of a speaker. |