One of the immediate challenges that the COVID-19 pandemic threw up was the challenge of maintaining the education system, as social interaction was almost impossible. It then becomes an urgent task to device a means to continue educating the young people.The most viable means was to use technologies to solve this problem. To be precise, Machine Learning (ML) and Artificial Intelligence (AI) came beckoning.
There is no doubt that Artificial Intelligence (AI) will be the dominant technology of the future and will impact every aspect of the human civilisation. Most importantly, AI/ML (machine learning) will influence how communication networks, a lifeline of our society, will be run.
And more so, when there is also the need to bring to its barest minimum communication barrier in diverse cultures and languages as found in Nigeria and Africa at large, a reason Nigerian Experts in the field of Information and communication technology (ICT) are proposing the: “AI‑Based Classroom” project. With AI‑based Natural Language Processing (NLP), classroom conversations between pupils and teachers are processed at the network edge to extract keywords while maintaining speaker anonymity.
These keywords are then transmitted to a trained classifier in a central server, which is able to recommend captivating media content based on the input, providing students with intuitive examples, supporting a teacher’s explanations. The media content can then be shared on a digital display in the classroom or used in any other format.
Though, rather than attempting to replace the efforts of elementary school teachers, the proposed system is designed to augment it, maximizing both school hours and off-school hours in the learning process. Regrettably, development of AI-based classrooms depends on efficient speech recognition libraries as a critical prerequisite. This proved very difficult to find.
A group of students are driving one of such projects, under the guidance of Dr. James Agajo, Associate Professor and Head of Wireless Networks and Embedded Systems Technologies (WINEST) Research Group in the Department of Computer Engineering at the Federal University of Technology in Minna, Nigeria.
WINEST Research Group launched a study on “use cases and solutions for migrating to IMT‑2020/5G networks in emerging markets” and the focus was to determine how machine learning could help emerging markets to leapfrog technology generations to take advantage of emerging and future networks while optimizing energy consumption, network coverage, and communication overheads.
Dr. Agajo said, “What we did was to come up with Automatic Speech Recognition as it relates to African Languages. You know ordinarily, this thing exists but in foreign languages. When you go to a high profile international meeting, you see people using a speech translator, language translator in their ear. When they are speaking French, the other person is reading it to you in English.
“And we are saying that why do we need to always surrender to technologies from the west? Why not come down and equally do our own so that there would be no barrier? If you are speaking Hausa, the other person speaking; who probably does not have an interpreter can easily use this ASR, that is an Automatic Speech Recognition mechanism we are putting together to hear what you are saying. So it has to break that barrier and at the same time lessen that gap.
“But to achieve this, one needs a database for all the speeches because how the stuff works is that; for every term you have in a particular language, English has a lot of words, those words are actually kept in a database; so that if you have Hausa, Yoruba, English, Idoma, whichever you are going to have, you can have all the words in a particular database.
“So, if you are able to capture all of the words in a database, then you could come up with a high level sophisticated program that would be selecting the item from a database and matching it with your own.
For instance, when you say in Yoruba, “wa”, then Hausa says “zo”, so you have the Yoruba word in the database, the Hausa word in the database too. When you speak each of these words in a language, it would check through the database and match it with an equivalent in a language of your choice.
That is why an issue of language barrier is not an issue in countries like the United States of America.
Dr James Agajo that has presented papers in high profile Conferences Organised by IEEE, Institute of Engineering Technology (IET), ITU, United States Environmental Protection Agency, National Space Research Development Agency, also said that: “We were in search of a speech recognition library which was able to function locally, meet users’ privacy concerns, and was freely available. In view of the extraordinary number of languages spoken across Nigeria, and Africa at large, we also needed a library able to perform well in processing the English language accented in many different ways.
“We evaluated many software libraries, but none of them succeeded in meeting all of these requirements. This led to the launch of a new WINEST Research Group project in February 2020 to develop a new speech recognition framework able to meet the unique requirements of the AI‑Based Classroom project.
“The project evolved from the discussions sparked by our presentation of the AI‑Based Classroom project at the 7th Regional Workshop on “Standardization of future networks towards building a better-connected Africa” in Abuja, Nigeria, 3–4 February 2020, convened by ITU’s standardization expert group for future networks and cloud computing, ITU Telecommunication Standardization Sector (ITU–T) Study Group 13.
“The expert feedback provided by the Abuja workshop motivated our launch of a pilot project in Nigeria to develop African Automatic Speech Recognition System (AASRS)”.
“We are collecting speech data and developing the ASR engine to deliver a prototype able to guide the development of a system ready for market deployment. We have developed the “Wazobia” mobile application, to support the necessary data collection, where Nigerian “voice donors” read displayed text aloud and donate the recording anonymously”. Agajo further revealed.
“Wazobia” is an amalgam of three words meaning “come” in Yoruba (wa), Hausa (zo) and Igbo (bia), Nigeria’s three largest linguistics groups.
Agajo also said that, “The speech data is stored on our server as “invalidated” by default, pending the crowd-sourced validation of this data by volunteers via the Wazobia mobile app. This validation results in Boolean evaluations of the accuracy of the ASR engine’s transcriptions of recorded speech.
To date, the project through the freely available Wazobia mobile app has collected over three hours of speech corpus from over 170 voice donors. This is a welcome development.
It is almost impossible to believe that the majority of the whole world would be on its knees begging for the Pfizer vaccine for the COVID-19; a Coronavirus that, 15 months ago little was known about. Africa should be proactive in developing solutions for Africans and the whole world.
- Prince Abdulrahman Obaje is a Media and ICT Consultant, Journalist, online marketer, social media strategist, Mathematician and Computer Scientist based in Abuja, Nigeria. He is the Founder and the Publisher of The Informavores!. You can reach me on +234 805 939 5252 or send i-witness report directly to me on firstname.lastname@example.org.
- Business and Economy2021.02.23FG Flag-off Gold Processing Cluster in Kogi
- News2021.02.22NGIJ-ABUJA Commends President Buhari Over the Appointment of Abdulrasheed Bawa as Substantive EFCC Boss
- ICT2021.02.22Improving Education In Africa Via ML, AI
- News2021.02.22Kogi State University Alumni Association (Worldwide) Appreciates Gov. Bello