I've got a problem with Google's MLKit face detector, as it returns a face even if the face is half-covered by something and this makes the face recognition model I use to think it's a new face, so I would like to know a solution for this problem, maybe a different face detector or another solution using MLKit face detector.
Thanks in advance.
ML kit works on data. Your results will be more accurate as much as more data you will provide to your model.If its taking half covered image it will also be beneficial for your trained model. You can train model by providing many images like with eye close, eyes open, left face, right face, look down, look up, zoom or half face covered etc.Once your model have enough data then it will recognized your even if you are wearing a face mask or even if your eyes were closed.
According to my opinion MLKit is much enough to implement face detection in your app. They are also improving it contineously. Happy Coding :)
