In the current sprint, one of my tasks was to use openCV to successfully open my computer’s webcam and take a photo. This is done, so I have been looking into the next step, which is using webcam photos and the python face_recognition library as facial recognition login for our chat app.

There are many tutorials for this task on the internet, and the two required libraries allow it all to be done without too many lines of code. Side note: I’ve looked at several of these tutorials, and they all use the same few public figures as examples for face recognition: Jeff Bezos, Elon Musk, Barack Obama, Joe Biden, etc. The lack of imagination is sad. Use someone unexpected! How about Steve Buscemi.

One potential problem I encountered with the photos my webcam takes is the poor image quality. I’m using a new Mac for the first time after running Windows my whole life. I had been led to believe Macs are better at this kind of thing. My office is relatively well-lit but I guess I need to plug in my ring light. I need to write the code to actually analyze a face to see how big of a problem this is.

Another problem: Looking over the face_recognition GitHub page, I noticed that it is technically *not* supported on Windows, only Linux and MacOS. There are work-arounds, which is somewhat comforting. I need to find out how that will affect our desktop app. I think (?) this means that any windows user who wants to use our app (including my group partner!) will have to follow somewhat complicated instructions to install the face_recognition library before being able to use the app. Or perhaps I’m totally wrong and there’s something about executable files that I don’t understand yet (since our app is a desktop app, we will have to turn in an executable).

These potential pitfalls of adding facial recognition login to our project are par for the course and a normal part of software development. I will have plenty of time to figure out solutions if I move quickly to write my first draft of code. So off I go to do that…