How to Solve False Facial Recognition for a Specific User?

FALSE ACCEPTANCE – Troubleshooting Guide

Please refer to the example below on how to set an individual score threshold value after checking the device’s system log file for a user’s score and solving the problem.

1. Normal/Correct Access The following log entries show that Adam’s (User ID:123) face was correctly recognized, with a facial threshold score of 3000:

Feb 21 09:50:56 ACTA4 local1.notice main[1969]: FACE-INFO: 214 score 3000.000000 

Feb 21 09:50:56 ACTA4 local1.notice main[1969]: FACE-INFO: id 214 pass

Feb 21 09:50:56 ACTA4 local1.notice main[1969]: last rcdid = 7

Feb 21 09:50:56 ACTA4 local1.notice main[1969]: AddEventLog(): rcdid=7, employeeID=214, createdate=1776736256, datetime=1776736256,  terminalSN=00111DB00011, event=1, Remark=FACE

2. False/Wrong Access On another device, Adam’s face was incorrectly recognized as John (User ID:234), with a facial threshold score of 5000:

Feb 21 10:30:56 ACTA4 local1.notice main[1989]: FACE-INFO: 315 score 5000.000000 

Feb 21 10:30:56 ACTA4 local1.notice main[1989]: FACE-INFO: id 315 pass

Feb 21 10:30:56 ACTA4 local1.notice main[1969]: last rcdid = 9

Feb 21 10:30:56 ACTA4 local1.notice main[1969]: AddEventLog(): rcdid=9, employeeID=315, createdate=1776736926, datetime=1776736926,  terminalSN=00111DB00011, event=1, Remark=FACE

3. Recommended Action

From this scenario, Adam’s individual score threshold should be set lower than 5000 (e.g., 4000) to avoid false acceptance.

Steps:

1. Log in to the device’s [View User List] page.

2. Click on Adam’s User ID 123.

3. Set the individual score threshold value to 4000.

4. Click [Modify].

5. Ask Adam to access the device again to verify correct recognition.