Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.
| Paper: | SP-L1.2 |
| Session: | Voice Conversion and Morphing Algorithms for TTS Systems |
| Time: | Tuesday, May 18, 15:50 - 16:10 |
| Presentation: |
Lecture |
| Topic: |
Speech Processing: Speech Synthesis (including TTS) |
| Title: |
SPEAKING STYLE ADAPTATION USING CONTEXT CLUSTERING DECISION TREE FOR HMM-BASED SPEECH SYNTHESIS |
| Authors: |
Junichi Yamagishi; Tokyo Institute of Technology | | |
| | Makoto Tachibana; Tokyo Institute of Technology | | |
| | Takashi Masuko; Tokyo Institute of Technology | | |
| | Takao Kobayashi; Tokyo Institute of Technology | | |
| Abstract: |
This paper describes an MLLR-based speaking style adaptation technique for HMM-based speech synthesis. Since speaking styles and emotional expressions are characterized by many segment-based features as well as frame-based features, it is necessary to adapt segment-based features for speaking style adaptation. To achieve segment-based feature adaptation, we utilize context clustering decision trees, which are constructed in the training stage, for tying of regression matrices. Using this technique, we adapt an initial ``reading'' style model to ``joyful'' or ``sad'' styles. Experimental results show that, using 50 adaptation sentences, speech samples generated from adapted models were judged to be similar to the target speaking styles at rates of 92% and 70% for joyful and sad styles, respectively. |
| |
| Back | |