Generation of Affect in Synthesized Speech

Sept. 1, 1989

Janet E. Cahn


When compared to human speech, synthesized speech is distinguished by insucient intelligibility, inappropriate prosody and inadequate expressiveness. These are serious drawbacks for conversational computer systems. Intelligibility is basic | intelligible phonemes are necessary for word recognition. Prosody | intonation (melody) and rhythm | clari es syntax and semantics and aids in discourse ow control. Expressiveness, or a ect, provides information about the speaker's mental state and intent beyond that revealed by word content

Related Content