Ghana NLP releases updates to language translators

Ghana Natural Language Process, Ghana NLP for short, has released advanced updates to their flagship applications; Khaya, the text-to-speech (TTS) and automatic speech recognition systems (ASR). The updates will upgrade user experiences and offer businesses and individuals the opportunity to do things they never could have done with any other existing applications.

Announcing the new features, Dr Paul Azunre, Founder & Director of Research at Ghana NLP shared short clips of the updates in action. The updates have enhanced the automatic speech recognition  capabilities previously offered by version.

Ghana NLP and Algorine which have worked together to deliver these tools to enhance the experience of Africans in these areas have added the updates to the Khaya app as well as the web translator. There are also application programming interfaces (APIs) for developers and businesses who want to integrate the ASR, text-to-speech and language translation capabilities to their projects.

For those using the Khaya app on Android or iOS, the updates are also available but only for pro users. With as low as GH₵6 (US$1) a month or GH₵60 (US$10) a year one can access these awesome features. This is specially very useful for content creators, musicians and movie producers. With the automatic speech recognition system, users can automatically extract text from videos in multiple languages including English, Ga, Twi, Ewe, Dagbani, Yoruba and Kikuyu.

About Ghana NLP

Ghana NLP is an open source project led Dr Paul Azunre, an MIT-trained Engineer and Computer Scientist. Comprised of a team of researcher, software engineers, data scientists and machine learning engineers among others, Ghana NLP’s primary focus is the natural language processing of Ghanaian languages and its application to local problems.

While their immediate focus was establishing roots in Ghana by building state-of-the-art NLP techniques for Ghanaian languages, they have since crossed the borders to include Nigeria and East Africa. As at the time of writing this report, their mobile application has support for 11 African languages and English. Of the 11, Dagbani, Ewe, Fante, Ga, Gurune, Kusaal and Twi are from Ghana while the rest cover other African countries including Nigeria where there is support for Yoruba.

Kikuyu which is spoken by at least 17.13% of Kenyans, the largest ethnic group in the country is also supported. Luo, spoken by at least 4 million people across Kenya, Tanzania, Uganda, South Sudan, Sudan, and northeastern Congo has support as well as Kimeru, spoken by the Meru people of Tanzania and Kenya.

At present, the ASR only supports six of the 11 African languages and English but with the team’s dedication and consistency, it will not be long time before more languages are added.

If you are enthusiastic about supporting technology in Africa and wants to support the work of Ghana NLP and their partner, Algorine, you can access their applications on both mobile and web via via Ghana NLP applications.