Method for identifying artificially generated speech, involves identifying automatic synthesis for classifying characteristics obtained from speech or forming speech signal which differentiates human voice and artificially generated speech

摘要

The method involves identifying an automatic synthesis for classifying characteristics (M) obtained from speech or forming a speech signal (S). The speech signal is provided for differentiating the human voice and the artificially generated speech. The prosodic characteristics are used for differentiation between human voice and artificially generated speech.