Khaya was the first translation app for Ghanaian languages relying on Machine Learning/Artificial Intelligence. Since version 1.0.2, it has started expanding to other African languages, such as Yoruba.
It also now includes Automatic Speech Recognition capabilities for Twi and Yoruba - speak directly to your phone and the app will transcribe your speech to text! For these languages, to the best of our knowledge, it is the first and only app to date offering such general speech recognition capabilities!
If you find the app useful and would like to see it grow and improve, kindly purchase an ad-free subscription to support our research. Thank you for your support!
The underlying algorithms are various neural network architectures, built by Algorine Research on top of open source technology developed by Ghana NLP.
If you encounter service outages, please report them using our developer website!
Features:
1. Speak to the app in Twi and Yoruba and it will transcribe it into text
2. Translate from English to a set of African Languages
3. Translate from a set of African Languages to English
4. Currently supported African Languages for text translation include Twi, Ewe, Yoruba and Ga (many more will be added soon!)
5. Submit improved translation corrections and recommendations directly from the app, and watch the technology improve to serve the African community
NOTE:: SYSTEM REQUIREMENTS
----------------------------------------------------
The app was built for Android 11. Some users with Android 7 have reported a good experience. Version 1.0.1 received positive feedback from some Android 5 & 6 users. The older the version of Android is, the more likely you are to experience problems however, especially with the latest audio-enabled release. Please report any such issues and we will do what we can to optimize for these situations, thank you!
Automatic speech recognition for Twi and Yoruba added, Yoruba text translations added. Text translations for Twi and Ewe have been significantly improved! Terms of service now prominently accessible within app.