Learn about automatic speech recognition in this 40-minute lecture from CMU's Multilingual Natural Language Processing course. Explore pronunciation modeling, acoustic modeling, and language modeling. Dive into topics like voice dialing systems, dynamic time warping, template matching, and training acoustic models. Understand how to estimate language model costs, measure ASR success, and discuss key points in speech recognition technology. Access additional course materials through the provided class website.
Overview
Syllabus
Automatic Speech Recognition
Voice Dialing System
Matching in Frequency Domain
Dynamic Time Warping
DTW algorithm
Matching Templates
DTW issues
More reliable matching
More reliable distances
Extending template model
Training an acoustic model
Language Model Estimate cost of sequence of words in the language • Need appropriate training data
Pronunciation Model
Measuring ASR Success
How good is good?
ASR Discussion Point
Taught by
Graham Neubig
