An Interview With Dr Ali Maina: A Nigerian Scientist Who Developed An Algorithm For Finding Missing People
19 February 2018
Last year, news was all over the UK about a new algorithm which could help in finding missing people, developed by scientists at the University of Bradford in England. The development was reported by the CNN, BBC, Telegraph, among others. We caught up with the scientist, Dr Ali Bukar Maina, a Nigerian, whose PhD work under Professor Hassan Ugail, led to the exciting innovation. Ali was kind enough to share the following information with us.
Tell us about yourself and schools attended
My name is Ali Bukar Maina, from Nguru, Yobe State. I attended Girgiri Primary School Nguru, then Suleja Academy, Niger State (now Federal Government Academy, Suleja) between 1992 and 1998, then did my bachelor's in Electrical and Electronics Engineering at the University of Maiduguri, Nigeria. Thereafter, I did a Diploma in Computing at Jigawa State Institute of Information Technology, an MSc in Software Technology with Network Management at the Robert Gordon University Aberdeen, Scotland before joining the University of Bradford for a PhD.
What is your area of research?
Computer vision and machine learning
Last year, your story was on CNN and BBC on a facial age progression algorithm you developed, can you tell us about it?
The algorithm was part of my PhD which entails teaching the computer to generate facial images of people at different ages. In simple terms, you feed in a picture, let’s say of a 30-year-old person, then the algorithm process it and generates an image of the person at any finite age(e.g. age 50). Alternatively, the algorithm can synthesise that image to a younger age (for instance back to how the person looked at age 20. This has several real-world applications such as the search for missing people. Indeed, the news on BBC and other news media reported our findings with regards to the image processing of Ben Needham, a British toddler who went missing over 20 years ago, till date he has not been found. Our algorithm generated some images of how he may look after all the years he was missing, which would thus make the search efforts easier.
Is this algorithm open source?
No, the algorithm is not open source. We hope to get a patent for it shortly.
Has this work been published in a scientific journal?
Yes, it has been published in the Journal of Forensic Sciences.
Ali M. Bukar, Hassan Ugail. Facial Age Synthesis Using Sparse Partial Least Squares (The Case of Ben Needham). Journal of Forensic Sciences, 2017; DOI: 10.1111/1556-4029.13523
Other improvements to the initial work have further being published as book chapters and conference articles.
What do you think are the other applications of this algorithm?
In addition to finding missing people, the algorithm can be used in the search for fugitives. For instance, in the next 10 or more years, most of the wanted criminals that couldn’t be caught will have changed in appearance. The algorithm could track their age progression and generate pictures that could identify them, which could lead to identification and arrests.
Can this type of research be done in Nigeria?
Computer vision and machine learning can be done in Nigeria. What is required is dedication, in-depth understanding of mathematics and programming. Furthermore, to teach (train) the algorithm, one needs a large database of labelled pictures and a powerful computer that would be able to do the training. These are things attainable in Nigeria.
How do you feel developing such a useful tool?
It gave me a sense of accomplishment seeing how mathematical equations that many people could not understand generated images of a person that people could be recognised.
Can you remember how you developed an interest in science?
I developed an interest in science when I was little, thanks to my dad, who has always been my motivator and mentor. He had a book called “Simple Science Experiments” which we used to study along with my siblings and try the experiments. That gave me a strong passion for science and motivated me to pursue it as a career.
Who is your science hero?
Geoffrey Hinton is a scientist I have high respect for. He is one of the most famous researchers in the field of Artificial Intelligence. His work helped kick off the world of deep learning (neural networks). He persisted in what he loved for decades when the whole world thought artificial neural networks made no sense. Today, all the tech giants, like Google, Facebook, Microsoft and their likes are using deep learning to achieve wonders.
What is your advice to young students and junior colleagues?
Hard work, dedication and perseverance are some of the ingredients for success. Rather than studying what will get you a job, pursue the subject you love
Further story on Dr Maina’s work can be found via:
Copyright statement: You may reproduce, republish, in whole or in part, this article. However, you should credit Science Communication Hub Nigeria and provide a link that links your readers back to this page.