Face Recognition Settings

Face Recognition Settings allow you to configure how facial recognition is used for worker clock-ins and attendance verification. These settings enhance security and support accurate time tracking across your organization.

Step 1: Go to the Fareclock Admin Console and go to Settings. Under ‘Settings’ find and click ‘Devices’.

Step 2: Click the ‘Defaults’ tab above.

In the Face Settings section:

• Allow employee to submit online face verify failure to manager for approval - If enabled, workers can submit a face verification failure request for admin approval when clocking in or out.
   
• Save punch if offline face recognition failure - When the time clock is offline, workers will be required to enter their PIN and take a facial photo when clocking in or out. Once the device goes back online, the punch is uploaded and the face recognition check is performed.

If face verification fails and this setting is enabled, the punch will still be saved but marked as an exception for admin review.

If this setting is disabled, the punch will not be saved. Instead, a Clock Log will be created for admin review, where it can be approved and converted into a punch if valid.

• Maximum number of photos in face model before face recognition may fail - If not enabled, or is disabled for a specific device, this setting determines the minimum number of photos required in a user’s face model before a failed verification blocks clock-in or clock-out.

This is especially useful for new employees or during initial setup when using Fareclock.

Face liveness detection strategies

• Eye or mouth movement - Requires the user to blink their eyes or move their mouth to pass face recognition verification.
   
• Face not on phone or photo - Requires the face to be presented directly to the camera and not through a photo, screen, or other display. Requires Time Clock version 3.5 or later.
   
• Passive Local - Uses a custom-trained AI model that runs locally on the device to detect visual indicators of face spoofing attempts, helping prevent the use of photos, screens, or other non-live images during face verification. Requires Time Clock version 4.0 or later.
   
• Passive Backend - Checks for visual indicators that may suggest the use of a photo, video, or other spoofing attempt during face verification. This method has stricter quality requirements and may result in more false positives if lighting, camera quality, or image conditions are not ideal.
   
• Block punch after liveness fail - If enabled, workers will be blocked from clocking in or out when face liveness detects a potential spoofing attempt. If disabled, the punch will still be recorded, but it will be flagged as a potential spoof for administrator review. Requires Time Clock version 4.0 or later.
   
• Crop photo to face - The mobile app can crop photos to focus on the user's face, reducing or removing the background. This can help improve face visibility by effectively zooming in on the face. If Detect Passive Face Liveness is enabled, photo cropping will not be applied. Lenient Cropping removes excess space around the face, especially when photos are taken in landscape mode.
   
• Allowed cameras - Specify which device cameras can be used for face detection and QR code scanning. Applies to Mobile App version 2.13 and later.
   
• Default camera - Specify which camera will be used by default when starting face detection or QR code scanning. Applies to Mobile App version 2.13 and later.
   
• Face scanner theme - Select whether the face scanner uses the new oval-framed interface or the classic full-screen interface. Applies to Mobile App version 2.13 and later.