Transferability of the audiobook-based recipe to other Kurdish varieties

Determine whether the audiobook-based Central Kurdish text-to-speech development recipe transfers to Kurdish varieties such as Badini and Hawrami that may lack comparable audiobook speech resources.

Background

The reviewed release demonstrates three single-speaker Central Kurdish systems trained on read audiobook and studio material. The paper emphasizes that this result depends on a substantial prepared speech resource and that the covered speakers, varieties, and registers are narrowly documented.

Because Badini and Hawrami are distinct Kurdish varieties and may not possess comparable audiobook corpora, it is unresolved whether the demonstrated data and training recipe can be applied to them in practice. This question matters for extending speech technology beyond Central Kurdish without assuming that the same resource conditions exist.