Course Detail
CSE6505
Speech Recognition
3 Credit Hour Course
Intended For Level 0 Term 0 Students
Prerequisite:
Introduction: speech production, perception, acoustic-phonetics, and signal representation; Algorithmic aspects of speech recognition systems: pattern classification, search algorithms; Stochastic modeling, and language modeling techniques; Noisy channel model; Bayes, HMM, Forward and Viterbi algorithm; Feature extraction and acoustic modeling and evaluation; Recognition system design and implementation: source coding, template training, performance analysis; Comparison of approaches to speech recognition; Advanced techniques for acoustic-phonetic modeling; Robust speech recognition; Speaker adaptation; Processing paralinguistic information; Speech understanding and multimodal processing.