Facial recognition and identification methods

¶ … Head

Recognizing Faces

There is a region in the brain, called the fusiform face area or FFA, which is vital in recognizing and distinguishing faces (Andrews et al., 2010). Brain scientists have been acquiring an understanding of the mechanisms, which distinguish faces. A baby's brain processes faces at four months as distinct objects. Faces tell a baby a lot of things, such as the person's identity, gender, race, emotion and truthfulness. The loss of that ability to recognize and distinguish faces is called proposophagnosia or face blindness. It often follows a stroke or brain injury. A person who suffers from this damage, even with normal vision, cannot recognize the faces he already knows very well, such as spouse and children. The damage is often on the fusiform gyrus, which is located in the underside of the brain and other areas (Andrews et al.).

Face Processing and Other Discoveries

A lot has been revealed by science about the general cognitive process involved in this function (Andrews et al., 2010). Faces are holistically processed rather than in parts. Focus is not made only on specific face parts, such as the eyes, nose or mouth. The entire information about these parts and their configuration are processed in creating a general model or concept of the face. This holistic procedure enables the brain to distinguish a specific face from many variations according to the individual encounter. A face recognition software today is not able to match the inherent capability of the human brain in this feat (Andrews et al.).

Imaging techniques have helped scientists more discoveries. One such discovery was that face-processing areas in the brain interact with memory networks (Andrews et al., 2010). This interaction enables the person to identify the one he is talking with or recall something of the past with this other person. Emotion has a particularly important role too. Some brain scientists believe that a defective connection with some emotional centers can induce the Capgras syndrome. This disorder enables the recognition of loved ones but that impostors have replaced them. Experience is likewise important. A child commits more face recognition misses than adults. And research found that face processing is like language in that faces less often recognized are more easily forgotten (Andrews et al.).

Face Recognition, Identification and Classification

Basic-level categorization is the common entry-point recognition for faces but not enough to identify them (Tarr & Cheng, 2003). The entry point is at the individual level. It would be later understood that faces can be recognized only holistically as opposed to the parts of an object. An object is recognized and discriminated according to their parts. Recent evidence shows that the fusiform face area lights up when objects are viewed in the same way as faces. It was originally thought that the area lights up only when faces are recognized (Tarr & Cheng).

Long-Term Memory and Face Recognition

Encoding and the retrieval process are essential to face recognition (Tarr & Cheng, 2003). The encoding process inputs the face features in long-term memory and helps in storing images permanently as well as for ease of retrieval later. The encoding must be correct or it can be lost. The stored information becomes like a databank, which is later used in recognizing or distinguishing faces. Encoding consists of a view-centered description process and an expression-independent description process. The view-centered description process is later used to represent the person's primary facial features or information, such as contrast or light intensity. This is influenced by the representation of the face, which includes the size, orientation, and the intensity of light. The expression-independent description process, on the other hand, is the abstract representation of the person's facial features and then the whole face for facial recognition. The description derives from the common characteristics of the face as viewed from different angles. The expression-independent description process compares codes sent from this description with another stored in the face recognition unit. If they match, the face is recognized. This kind of encoding helps in the matching and in retrieving long-term memory and facial recognition. Encoded information must not be distorted and must be located for easy retrieval (Tarr & Cheng).

A study asked 10 healthy young adults, 8 men and 2 women, to perform face encoding, face recognition, face perception control and sensorimotor control (Haxby et al., 1996). These were measured in the regional cerebral blood flow with the PET equipment. The experiment found no overlap between the brain structure in these memory functions. Findings also showed that encoding activated the left prefrontal cortex while recognition activated the right prefrontal cortex. Overall findings revealed that the hippocampus and the adjacent cortex become involved in memory function at the time of the encoding of new memory. Face recognition was shown to involve anatomically dissociable operations rather than simply by recapitulation (Haxby et al.).

Error in Face Recognition

Neuropsychological investigations report that false self-recognition or mis-identification of faces can result from massive damage to the right hemisphere (Rapcsak et al., 1994). Extreme reliance on the feature-based, left-hemisphere approach to face processing and a damaged face recognition system's inability to critically evaluate output result in recognition errors and mis-identification. The findings of these investigations suggest that the feature-based left-hemisphere recognition system is prone to error. Its component facial features are often common to many different persons. This is so also because the reliable recognition and identification of faces is heavily dependent on the efficient processing of the configuration information of faces by the right hemisphere. Researchers propose that decision-making and monitoring functions of the face recognition system be more confined to the right frontal lobe (Rapcsak et al.).