Responsible AI

Developing predictive models for good intentions to the community is something that should be considered vital. The growth of Machine Learning and Artificial Intelligence has led to rise of sophisticated technologies such as facial recognition. Facial recognition is being used on different phenomena, from security, to other service provisions.

However, this technology is far from perfection since it is a lucid fact that it performs poorly when it comes to people of color (African-American). Different researches has shown that most of the widely used facial recognition models return false positive on black people, especially black females.

Taking this to consideration, some metropolitan units use facial recognition for criminal monitoring and tracking. As these models return false positive on people of color, may result to suing innocents predicted as criminals. This violates the principles of Responsible AI.

Responsible AI is a modern concept due to problems that arose with the growth of AI. Problems like misidentifying civilians as criminals through facial recognition, may cause life-long effects to victims. Speaking ethically, this issue rises concerns about racism and… you know what that will lead to.

Some researchers say this may be caused by the fact that most of the majority of data used to train these models contains more white faces than black faces. Here there’s a research to do. And as People of Color, we have take this as a research field and not raising racism pressure on social media across the internet.

Some say, it is hard to capture a quality photo of a black person. This leads to algorithms fail to extract as much features as possible on black faces as it does on white faces. I had conversations with experts in photography and photo editing and they agreed that

So, we can see that this technology is still far from perfection and it should not be used on use cases that involve subjecting guilty, as there are still years of research before having full trust and hope on facial recognition technology.

We, African AI researchers, we should take this as a research area and try to do our best to solve or at least contribute upon eradicating this issue. And be ambassadors of telling people that it has got nothing to do with racism, ethnicity or anything. Rather, it is still a growing technology that is taking its baby steps towards perfection.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store