Pronunciation Modeling of Mandarin Casual Speech

7/26/00

Click here to start

Table of Contents

Pronunciation Modeling of Mandarin Casual Speech

Baseline Experiments (1 week)

Generalized Initial/Final (GIF) Phoneme Models (2 weeks)

Probabilistic Pronunciation Lexicon using GIFs(2 weeks)

Other Experiments (1 week)

MPM Baseline Acoustic and Language Models (Bill Byrne)

MPM Baseline Acoustic and Language Models-Broadcast News System (Bill Byrne)

MPM Baseline Acoustic and Language Models (Bill Byrne)

Chinese Annotated Spontaneous Speech (CASS) Corpus

Chinese Annotated Spontaneous Speech (CASS) Corpus

Generalized Initial/Final (GIF)

Generalized Syllable (GS)

Output and Transition Probabilities of GIFs (SZJ)

GIFs: Initial Experimental Results (Liu Yi)

Pronunciation Modeling using Decision Trees (Veera)

A Decision Tree

Discussions with Linguist

Alignments between canonical and annotated phones

Adding Voicing to the Decision Tree (Terri Kamm)

Adding Voicing to the Decision Tree (Terri Kamm)

Syllables (Initial/Final) Alignment Program (Umar)

End of Presentation

Examples: Sound Changes vs. GIFs

Examples: Output & Transition Probabilities