Liveness Validation Selfie Guide on Talenta Live Attendance

Article author
Learning Center Mekari
  • Updated

When doing Live Attendance on Talenta Mobile, a selfie that meets certain criteria is needed to do online attendance. One of the features that can detect the criteria for certain living things is Liveness Validation on Talenta. This process is useful for ensuring that there is no fraud when employees take attendance directly by detecting whether the selfie taken is a real face or not.

Liveness validation uses the face occlusion function which can assess the structure and feature algorithms on the face, such as the following facial features.
Faceocclusion__6_.png

Therefore, if an object is found that is blocking that part of the face, then the selfie results of your face will not be identified optimally and will most likely be accepted by the system.

The following is an example of a selfie case that will be accepted and rejected by the system.

A. Examples of comparison cases accepted and rejected

No.

Case

Accepted

Rejected

1

Wearing Hijab

Faceocclusion__1_.png

All facial features such as forehead, eyebrows, pair of eyes, nose, right and left cheeks, and mouth are clearly visible.

Faceocclusion__8_.png

Facial features such as the forehead and eyebrows are covered by the hijab.

2

Glasses

Faceocclusion__11_.png
The glasses worn do not block facial features such as eyes, eyebrows, and cheeks. In addition, the lighting in the room is clear.

Faceocclusion__3_.png
Glasses worn indoors in low light cause reflections from the mobile phone screen to cover the eyes. In addition, the eyebrows are not clearly visible.

3

Hair on the forehead

Faceocclusion__4_.png
Even though half of one eye and eyebrow are covered by hair, this type of selfie can still be recognized by our system, although it is best avoided.

Faceocclusion__7_.png
Facial features such as the forehead and eyebrows are completely covered by hair. 

4

Case with face obstructed by object

Faceocclusion__9_.png
A small part of one cheek is covered by one finger, but the other facial features are still clearly visible. 

Faceocclusion__5_.png
One side of the face was completely covered by the hand.

5

Facing straight ahead of the camera


Face facing straight ahead of the camera with good lighting.


The face is not facing straight at the camera.

6

Face orientation


Portrait  (vertical) face orientation .


Landscape (horizontal) face orientation

7

Face position


Position your face right in the middle of the camera's coverage.


Position the face in the camera's field of view.

8

Background


Plain background.


The background is crowded and full of objects.

B. Rejected case examples

No. Case Rejected 
1 Stylish glasses case and  face paint/tattoo Faceocclusion__2_.png
Case of face with  face paint  or  tattoo.
Faceocclusion__2_.jpg
The upper facial features are covered by reflective or dark-style glasses.
2 Cases of using face masks Faceocclusion__10_.png
Wearing a face or mouth pattern mask.
Faceocclusion__1_.jpg
The lower facial features are covered by a mask.
3 Bad lighting case
The room light is too bright.

The room lighting is too dim.

The room lighting is colorful and unnatural. For example: Red.

Turn your back to the spotlight or the sun.
4 Accessories
Even though you wear clear lens glasses, the lenses reflect light from  your cellphone , causing the shape of your eyes to become distorted.

Wear a hat that covers the forehead and part of the face.
5 Blurred/ blurry face photo
The photo is unclear because the object is moving, resulting in a blurry and distorted image.

The photo is unclear because the object is moving, resulting in a blurry and distorted image.
6 Photo with Identity Card
Avoid taking photos with your ID card because the system will detect objects on the card.

Avoid taking photos with your ID card because the system will detect objects on the card.

This is the Liveness Validation selfie guide on Talenta Mobile. To learn more about how to use Liveness Validation in Live Attendance, read here.