Adaptive phonetic segmentation techniques in dysphonic voice
This task aims at developing computational rules enabling to select those regions in the whispered speech or dysphonic voice that should be subject to artificial voicing.
The outcomes in this task, in addition to a task report, are algorithms implementing adaptive segmentation techniques that are required in whispered speech to voiced speech conversions (Task B.2).
Milestones
M3_b: Adaptive segmentation algorithms.
Other expected outcomes include a task report.
Leader: FEUP