This is Festvox voice trained on Arabic Speech Corpus
Note that training on high quality corpus like SASSC would give better results (samples included).
ASC corpus: Festvox sample1, Festvox sample2, Flite sample1 and Flite sample2
SASSC corpus: Festvox sample and Flite sample
sudo apt install build-essential
sudo apt install git
sudo apt install sox
sudo apt install libcurses-ocaml-dev
git clone https://github.com/asrajeh/arabic-tts.git
cd arabic-tts
./install_asc_voice.sh
./tts.sh -i input.txt -o output.wav -v asc_festvox
./tts.sh -i 'السَّلَامُ عَلَيْكُمْ' -o output.wav -v asc_festvox
./tts.sh -i 'السلام عليكم' -o output.wav -v sassc_festvox -d 1
./build_asc_voice.sh