Detecting human faces in a scene has been achieved with good success by many computer vision teams across the world. However, identifying an individual from the captured face-image is still considered a difficult task particularly if the face-image database is large. One of our solutions developed using our core expertise can identify an individual from his or her image from a large database in real-time. The unique capability of our system is to accommodate varying illumination conditions and views of the face-image that are slightly different from the training set contained the database.
A further advanced solution developed by our team can not only detect a human face but also detect the facial key-points such as eyes, nose, mouth, ears (if visible in the frontal view). Robust tracking algorithms have also been developed to track these facial key-points, once detected, in subsequent frames.
Market analytics (People counting, gender and age recognition)