Table of Contents
1. Introduction
1.1 Project Description
1.2 Introduction to Speech Synthesis
2. History and Development of Speech Synthesis
2.1 From Mechanical to Electrical Synthesis
2.2 Development of Electrical Synthesizers
2.3 History of Finnish Speech Synthesis
3. Phonetics and Theory of Speech Production
3.1 Representation and Analysis of Speech Signals
3.2 Speech Production
3.3 Phonetics
3.3.1 English Articulatory Phonetics
3.3.2 Finnish Articulatory Phonetics
4. Problems in Speech Synthesis
4.1 Text-to-Phonetic Conversion
4.1.1 Text preprocessing
4.1.2 Pronunciation
4.1.3 Prosody
4.2 Problems in Low Level Synthesis
4.3 Language Specific Problems and Features
5. Methods, Techniques, and Algorithms
5.1 Articulatory Synthesis
5.2 Formant Synthesis
5.3 Concatenative Synthesis
5.3.1 PSOLA Methods
5.3.2 Microphonemic Method
5.4 Linear Prediction based Methods
5.5 Sinusoidal Models
5.6 High-Level Synthesis
5.6.1 Text Preprocessing
5.6.2 Pronunciation
5.6.3 Prosody
5.7 Other Methods and Techniques
6. Applications of Synthetic Speech
6.1 Applications for the Blind
6.2 Applications for the Deafened and Vocally Handicapped
6.3 Educational Applications
6.4 Applications for Telecommunications and Multimedia
6.5 Other Applications and Future Directions
7.1 Speech Application Programming Interface
7.1.1 Control Tags
7.2 Internet Speech Markup Languages
7.3 MPEG-4 TTS
7.3.1 MPEG-4 TTS Bitstream
7.3.2 Structure of MPEG-4 TTS Decoder
7.3.3 Applications of MPEG-4 TTS
8. Audiovisual Speech Synthesis
8.1 Introduction and History
8.2 Techniques and Models
9. Products
9.1 Infovox
9.2 DECTalk
9.3 Bell Labs Text-to-Speech
9.4 Laureate
9.5 SoftVoice
9.6 CNET PSOLA
9.7 ORATOR
9.8 Eurovocs
9.9 Lernout & Hauspies
9.10 Apple Plain Talk
9.11 AcuVoice
9.12 CyberTalk
9.13 ETI Eloquence
9.14 Festival TTS System
9.15 ModelTalker
9.16 MBROLA
9.17 Whistler
9.18 NeuroTalker
9.19 Listen2
9.20 SPRUCE
9.21 HADIFIX
9.22 SVOX
9.23 SYNTE2 and SYNTE3
9.24 Timehouse Mikropuhe
9.25 Sanosse
9.26 Summary
10. Speech Quality and Evaluation
10.1 Segmental Evaluation Methods
10.1.1 Diagnostic Rhyme Test (DRT)
10.1.2 Modified Rhyme Test (MRT)
10.1.3 Diagnostic Medial Consonant Test (DMCT)
10.1.4 Standard Segmental Test
10.1.5 Cluster Identification Test (CLID)
10.1.6 Phonetically Balanced Word Lists (PB)
10.1.7 Nonsense words and Vowel-Consonant transitions
10.2 Sentence Level Tests
10.2.1 Harvard Psychoacoustic Sentences
10.2.2 Haskins Sentences
10.2.3 Semantically Unpredictable Sentences (SUS)
10.3 Comprehension tests
10.4 Prosody evaluation
10.5 Intelligibility of Proper Names
10.6 Overall Quality Evaluation
10.6.1 Mean Opinion Score (MOS)
10.6.2 Categorical Estimation (CE)
10.6.3 Pair Comparison (PC)
10.6.4 Magnitude and Ratio Estimation
10.7 Field Tests
10.8 Audiovisual Assessment
10.9 Summary
11. Conclusions and Future Strategies
Appendix A: Speech Synthesis Demonstrations
Appendix B: Summary of Speech Synthesis Products