Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

YouTube

Speech Recognition in the Context of Variability: Lessons from Human Listeners

Center for Language & Speech Processing(CLSP), JHU via YouTube

Overview

Explore the complexities of speech recognition and acoustic variability in this lecture by Dr. Shannon Heald from the University of Chicago. Delve into the challenges posed by acoustic-phonetic variability between talkers and learn how human listeners adapt to understand difficult speakers. Discover how leveraging systematic variance in speech signals allows listeners to maintain robust phonetic constancy despite noise or distortions. Examine traditional models of speech perception, the nature of talker variability, and the role of expectations and visual information in recognition. Gain insights into Dr. Heald's research on how listeners reorganize their processing of acoustic features when adapting to new talkers. Understand the implications of this work for developing more robust speech recognition systems that can handle real-world variability.

Syllabus

Introduction
Variability in Speech
Acoustic Variability
Traditional Models of Speech Perception
Omen and Colin 1989
Two Perspectives on Recognition
Understanding the Nature of Variability
Examining Talker Variability
F1 Variability
Sources of Variability
Main Takeaways
Citation Form Values
Understanding Variability
Variability and Recognition
Expectations alone
Visual information
Recognition as a perceptual learning process
Synthetic speech
Analysis
Conclusion

Taught by

Center for Language & Speech Processing(CLSP), JHU

Reviews

Start your review of Speech Recognition in the Context of Variability: Lessons from Human Listeners

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.