Prosody-based speech synthesis by unit selection for Arabic
Raja Abdelmalek, Zied Mnasri · 2020
This work aims to develop a high quality Arabic concatenative speech synthesis system based on unit selection. The original unit selection algorithm was modified to integrate more phonological, linguistic and contextual features in order to improve the selection cost calculation fromone side, and more prosodic parameters for more exact concatenation cost estimation from the other side. The objective and subjective assessments based respectively on SER (Signal-to-Error Ratio) and on MOS (Mean Opinion Score) tests show satisfactory results.