Kobixa — presentation audio credits

Original scripts, animated artwork and procedural music score prepared for Kobixa.
Music: original pad and arpeggio composition (92 BPM), no third-party recordings or samples.
Synthetic speech generated locally using Piper 1.8.0 (Open Home Foundation):
https://github.com/OHF-Voice/piper1-gpl (engine GPL-3.0; engine/models not distributed with this site).

French: fr_FR-siwis-medium, SIWIS dataset (CC BY 4.0).
Dataset: https://datashare.is.ed.ac.uk/handle/10283/2353
License: https://creativecommons.org/licenses/by/4.0/
Model card: https://huggingface.co/rhasspy/piper-voices/blob/main/fr/fr_FR/siwis/medium/MODEL_CARD
English: en_US-ljspeech-high, trained from scratch on LJSpeech (public-domain corpus).
https://keithito.com/LJ-Speech-Dataset/
https://huggingface.co/rhasspy/piper-voices/blob/main/en/en_US/ljspeech/high/MODEL_CARD
Arabic: ar_JO-kareem-medium.
Model card and dataset reference:
https://huggingface.co/rhasspy/piper-voices/blob/main/ar/ar_JO/kareem/medium/MODEL_CARD
https://github.com/AliMokhammad/arabicttstrain/

These are synthetic voices, not recordings of a spokesperson or a cloned user voice.
All synthesis, rendering and mixing took place on the development machine.