Adapting F5-TTS Model to Kurdish Sorani: Diffusion-based Speech Synthesis with a Specialized Dataset. UHD Journal of Science and Technology, v. 9, n. 2, p. 198–207, 6 Oct.2025.