Speech Recognition and Understanding

This is an introductory course on Speech Recognition and Understanding offered to undergraduate and graduate students. The goal is to develop understanding of traditional as well as contemporary speech processing techniques and enable students to develop applications using existing tools. Students will be given theory and programming assignments targeted towards solving real-world speech recognition problems. Groups of two or three will do a course project and show a demonstration at the end of the semester.

1. Describe fundamentals of speech production, distinguish between speaker dependent and independent components of speech and develop appropriate representations.
2. Model recognition problems using traditional (HMM+GMM) and advanced techniques (DNN) for automatic speech recognition (ASR).
3. Integrate language models with speech recognizers to develop specific and generic speech recognition applications.
4. Build Speech and Speaker Recognition Systems