Work Logs: Camera Notices During Face Verification
When Face Recognition is on and the camera does not give a usable picture, the day shows a notice in WORK LOGS reporting how many checks were affected and when. It means the camera may not have been available — not that anyone did anything wrong.
There are plenty of ordinary reasons: another app holding the camera, a closed laptop lid, a privacy shutter left across the lens, a driver that stopped working, or simply a broken webcam. If the employee was warned about it, the notice says so.
Treat it as a prompt to check the equipment. Ask first — it is nearly always a hardware or software problem.
One notice deserves special mention: “Camera malfunction”. It appears when the camera switched on but never sent any video — typically a stuck driver, a USB power glitch, or a failing webcam. In that case the employee also sees a popup on their own screen (at most once every 30 minutes) advising them to restart the computer, and your work log shows the notice at the same time.
You don’t need a technician for this one. Send the employee our step-by-step guide: Fix the “Camera malfunction” popup on Windows. It shows how to confirm the problem in one minute with the built-in Windows Camera app and walks through the usual fixes, from a simple restart to drivers — ending with a clear answer on whether the camera needs replacing. Once the camera works again, face checks resume on their own; nothing needs to be reset in SCREENish.
A different kind of notice: “Spoof suspected — camera feed may be fake”. The camera notices above mean the camera gave no usable picture. This one means the opposite: a face was in the picture, but across a sustained run of consecutive checks the frames kept failing the liveness test, the way a photo or a phone screen held up to the webcam does. A single failed check never raises it, and a covered lens, an empty chair or a dead camera never raise it either.
The card shows how many reports arrived that day, between which times, the longest streak of suspect checks, and the camera device when it is a virtual one. The condition clears on its own as soon as a live check verifies, and comes back if the pattern resumes.
Treat it the way you treat a low match rate: as a prompt to look and ask, not as a verdict. Open the hourly composite next to the notice and look at the frames of the failed checks. Honest setups fail the liveness test now and then: strong backlight from a window, heavy smoothing or beautify filters on the webcam, a very low-quality camera at night, or a phone used as a webcam through an app that re-encodes the picture. A virtual camera on its own is never treated as cheating; background-blur tools and conferencing software present themselves that way, which is why the device is shown as information only.
What the check catches, what it only flags and what it cannot see is set out on the liveness detection page.