Hidden Cues in Voice: Stress, Emotion, and Lombard Effect

ISBN-13: 9781614517153 / Angielski / Twarda / 2021 / 266 str.

John H. L. Hansen; Hynek Boril

Opis

The book provides a survey on variability of speech production due to stress, emotions, and Lombard effect, and its impacts on speech-enabled systems. The authors analyze characteristics of voice that are dominant in producing different speech modalities, as well as the role of speech variability in automatic speech processing and the techniques that can objectively assess and equalize such variability. "

Spis treści

Part A. Production and Perception of Speech Under Stress, Emotions, and Lombard Effect

1. Speech Production Variability
2. Models of Speech Production
3. Fundamental Frequency, Duration, and Intensity
4. Glottal Source and Vocal Tract Effects
5. Perception of Speech Variability
6. Auditory Models
7. Human Perception of Speech Modalities

Part B. Speech Variability and Automated Speech Systems

8. Why do Speech Systems Break?
9. Sources of Distortion and Mismatch
10. Speaker, Environment, and Recording/Data Capture Based Variability
11.Speech and Speaker Recognition
12. Speech Variability in Speech Recognition Performance
13. Speech Variability in Speaker Recognition Systems
14. Domains for Improving Speech & Speaker Recognition Systems
15. Need for Speech Assessment in Speech & Speaker Recognition
16. Automatic Speech Assessment
17. Can We do Better? Boosting Robustness
18. Feature Extraction
19. Acoustic Modeling
20. Normalization/Compensation Schemes
21. Future Directions

O autorze

John H.L. Hansen, UTD, Dallas, Texas; Hynek Boril,UTD, Dallas, Texas.