How To Use
This page contains some basic instructions around using the tool and understanding the feedback that you'll receive.
Can You See Me has been designed for people who don't like reading manuals. Once opened the tool will very briefly instruct you on how to use all of its features accordingly:
- Press enter to hear an approximation of where your face is (see below for an explanation of the feedback you will receive).
- Press d to receive a slightly more detailed approximation which includes your face’s location expressed as percentages (See the frequently asked questions page for information on how these are calculated).
- Press p to take a photo of what your webcam can currently see (the photo will be saved in the folder where Can You See Me is located).
- Press escape to exit.
Can You See Me divides the image your webcam can see (AKA the frame) into a 7 by 7 grid in an effort to provide the most useful feedback possible. The first piece of feedback you will receive represents where your face is positioned vertically and the second where your face is positioned horizontally.
- Vertical positions are at the bottom, towards the bottom, slightly towards the bottom, in the middle, slightly towards the top, towards the top and at the top.
- Horizontal positions are to the left, approaching the left, slightly to the left, in the middle, slightly to the right, approaching the right and to the right.
In practice, this means that you will receive pieces of feedback such as "towards the bottom and slightly to the left", "in the middle" (which means your face is roughly in the centre of the frame) and "at the top and to the right" to give some examples. If you find this confusing, I'd suggest that you experiment by moving your upper body around in relation to your webcam and seeing what feedback you get for different positions. You should however bear in mind that depending on where you are positioned in relation to your webcam and on how adjustable your webcam is it may not be possible to position your face in every single position of the grid.
If after the feedback Can You See Me announces "low confidence" or "very low confidence" it has detected multiple faces within the image, which means that (unless there are multiple people in shot) the technology is incorrectly classifying something that it's seeing as a face. In this instance there is unfortunately no way of telling which of the faces it has found is the correct one or not.
Additionally, when using Can You See Me, it is important to remember that you are being told where your face is and not where it should be, which is largely down to personal preference. Having said this, positions such as "at the top and in the middle" and "towards the top and in the middle" are usually sensible positions to aim for.