LING 1100 Lecture Notes - Lecture 11: Diphone, Articulatory Synthesis, Speech Synthesis
Document Summary
Is the study of computer processing, understanding, and generation of human languages. Human language systems and computers are both information processing systems. Modelling language computationally may give insights into how humans process language. Alan turing, sometimes referred to as the father of computer science and arti cial intelligence (worked on code breaking in ww2) Judge listens to computer and a human and has to decide which is the human. Intelligent web searching intelligent spell checking information retrieval from corpora. Di erent from speech recognition, although the terms are often used interchangeably. Discrete (words uttered separately) vs. continuous (normal speech) Small-vocabulary & many users/ large vocabulary & few users. Concatenative synthesis: stringing together segments of recorded speech. Diphone synthesis: uses a small recorded database containing all the diphones (transitions between two sounds) of a language. Unit selection synthesis: uses recordings of one or many of all possible types of units: sounds, diphones, syllables, words