On April 4, 2013, Lakshmi Saheer successfully defended her PhD thesis
On April 4, 2013, Lakshmi Saheer successfully defended her PhD thesis entitled "Unified Framework Of Feature Based Adaptation For Statistical Speech Synthesis And Recognition".
The work described in her thesis takes place in the context of personalized speech-to-speech translation project with the aim for rapid speaker adaptation. The work proposes vocal tract length normalization and its combination with other linear transformations in a unified framework for speech recognition and synthesis to bring in target speaker characteristics with little adaptation data of the order of few seconds.
Congratulations to her.
For more details, please download the thesis here:
Unified Framework Of Feature Based Adaptation For Statistical Speech Synthesis And Recognition