HMM-based Speech Synthesis:
Text-to-speech synthesis with
arbitrary speaker's voice

Sampling rates, Auditory scales, Generalized logF0
(Another samples)

Gap between ASR and HTS

Geographical GUI for
HMM-based Speech Synthesis

1000s voices trained on
various ASR corpora


Lecture by HTS-2007 voices

Robustness of HTS-2007

Application to many languages

Multi-dialectal HTS

Speaker interpolation and morphing

Adaptation to child speech
more samples (1) (2)

HSMM-based MLLR adaptation

Various speaker's voice

Speaker adaptive training &
Shared-tree-based clustering

CSMAPLR adaptation

LF model (improved source model)

Speech Synthesis with Various
Emotional Expressions &
Speaking Styles

Style modeling

Style adaptation

Style interpolation and morphing (English samples)

Style control

HMM-based Speech Production

Articulatory-controllable acoustic parameter generation
(more details)

Formant-cotrolled acoustic parameter generation

Applications of HTS

Speech-to-Speech Translation using HTS

Human walking motion synthesis

Speech driven head motion synthesis

Automatic speech-driven animation