Online Face Identification System Using Real-Time Eyes and Mouth Tracker

Researchers have built a system which can identify a person using their eyes and mouth as independent components of the identification process. Furthermore, its ability to track within ±45 degrees to right and left, and within ±15 degrees from frontal pose will foil any attempt to use a photo for identification purposes.

EMoFaceID: Online Face Identification System Using Real-Time Eyes and Mouth Tracker (EMoTracker)

EMoFaceID is a system that can identify a person using eyes and mouth, instead of using entire face area
(holistic method). Each of these features is utilized as an independent component in the identification
engine. As a result, the system can perform well even though the mouth is covered or temporal occlusion
occurred. EMoFaceID identifies a person passively, which means the person to be identified does not have
to be aware of the camera, and it also performs in real-time. Thus, it can be used in online identification
systems like web-based log-in applications or surveillance security systems. EMoFaceID starts by detecting
and tracking the face, eyes and mouth using EMoTracker (Eyes and Mouth Tracker). As EMoTracker can
perform tracking within ±45º to right and left, EMoFaceID can handle identification within ±15º from frontal
pose. This advantage can be the solution to any attempt to use photo of a person in front of the camera for
identification purposes.

For more information, please contact
Dr. Shahrel Azmin Suandi
(Contact details available to registered journalists from the link below)

Published: 14 Aug 2007

Contact details:

Universiti Sains Malaysia 11800 Minden Penang

++604-6533888 (Main Campus), ++609-7651704/00/11(Health Campus), ++604-5937788 (Engineering Campus)
Country: 
News topics: 
Content type: