CSCI-6302 (3) Speech Recognition and Synthesis

Introduction to automatic speech recognition and understanding, conversational agents, dialogue systems, and speech synthesis/text-to-speech. Topics include the noisy channel model, Hidden Markov Models, A* and Viterbi decoding, language modeling (N-grams, entropy), concatenative synthesis, text normalization, dialogue and conversation modeling. Recommended prereqs., CSCI 5582 or 5832, or LING 5200, or instructor consent required. Requisites: Restricted to graduate students only.