Utilising one-to-many feature based face recognition and matching from large and challenging face databases
VIFRS is our proprietary product focused on identifying an individual from a large database of images that have been used to train the system. The face recognition system not only assists in identifying the individual but also in matching several faces located within a particular database. It is capable of recognising and matching obscure images in the form of sketches or low resolution snapshots to images trained within the database and successfully returns matching images.
VIFRS can be applied to many situations including criminal investigation situations by matching from police databases for face based criminal or offender detection using sketches or snapshots. This can also narrow down potential suspects by providing a form of gender detection and similar facial feature detection. Additionally staff and customer recognition for entry security applications can be implemented by using the VIFRS as the backbone of the application.
With real-time database updating compatibility, VIFRS boasts a quick response time when recognizing distinct faces from a snapshot or approximate sketch and providing the closest match detected followed by other images with similar features. This has been tested against a large database of 10,000 images and has provided results with a very fast response time.