Khaya.AI, an artificial intelligence platform founded by Ghanaian researcher and entrepreneur Dr. Paul Azunre, has unveiled new speech-recognition and text-to-speech technology covering 32 African languages, a development the company says could digitally empower an estimated 540 million people across the continent.
In a video demonstration, Azunre, also known musically as Dr. Pushkin, showcased the system’s ability to accurately transcribe and interpret spoken language in real time. The platform allows users to upload or record audio directly in a browser and receive immediate transcription.
“So now I have it transcribe once you submit,” he said in the demo, walking users through examples in Adangme, Ewe, Fante, Ga, Nzema and Asante Twi. “If you compare the reference to the transcription, you will see that it’s doing pretty well.”
Dr. Azunre said the tools are intended to expand digital access for speakers of African languages that have historically been excluded from global language-technology development. Khaya.AI is currently the only platform covering all government-sponsored Ghanaian languages.
“Today I wanted to let you know that you can test our text-to-speech and our speech-recognition AI for 32 African languages, covering approximately 540 million people, which is more than half a billion,” he said.
The new tools, the company says, could help broaden market access for African businesses, strengthen digital communication across multilingual communities, and support education and documentation efforts for endangered languages.
Azunre, a graduate of MIT and founder of the Ghana-based AI research company Algorine, has spent recent years developing tools to increase linguistic inclusion in technology. He has previously led projects focused on machine learning applications in African languages through Khaya.AI.
Source: a1radioonline.com|101.1Mhz|Mark Kwasi Ahumah Smith|Bolgatanga
