Emotional forecast through CCTV video
IT IS nowadays usual in electronic media that we hear about video recording of criminals committed offense and successfully escaped from the crime scene. Such video footage remains an exclusive news highlight for the remaining hours of the day and then law enforcement agencies starts investigation in a conventional fashion asking different intelligence sources about the faces of the criminals. Surprisingly not all the time this story happens sometimes criminals are smarter than as estimated, they took away the DVR (Digital Video Recorder) along with them leaving no trial behind!! To cope with this challenging situation, a wiser solution is needed so as to handle law breakers with an iron hand. One such scientific solution is hidden in “Facial Action Coding System” (FACS).
Dr. Paul Ekman, an American psychologist, in 1957 first published his research work describing ways to empirically measure nonverbal behaviour of humans. He studied human emotions and their relation to facial expressions and after a years of cross cultural research he finally created an atlas of emotions with more 10,000 facial expressions. However, in 1950s there was a more popularized belief that facial expressions and their meanings are culturally determined through behavioral learning processes but later Paul Ekman through a series of studies found a high agreement among the Western and Eastern cultures that their emotional labels fits into a common facial expressions system. Facial expressions he found to be universal includes those indicating anger, disgust, fear, happiness, sadness, and surprise. To prove and further investigate into the matter Paul Ekman published his famous emotion recognition stimulus called “Picture of Facial Effect” (POFA). It consists of 110 black and white images of caucasian actors portraying the six universal emotions plus neutral expressions, the POFA has been used to study emotion recognition rates in normal and psychiatric populations around the world. Many researchers favor the POFA because these photographs have been rated by large normative groups in different cultures.
In 1978 Dr. Paul Ekman and Wallace V. Friesen developed the Facial Action Coding System (FACS) to classify every human facial expression. It is an anatomically based system for describing all observable facial movement for every emotion. In its simple term facial action coding system looks at facial expressions with respect to facial muscles. FACS describes facial movement on the basis of elemental units called “Action Units” or AUs, each AU represents the movement that occurs when underlying muscle fibers contract. The process involves identification of AUs based upon the location and/or type of action involved e.g. AUs in the upper face which affect the eyebrows, forehead, and eyelids. Then the lower face AUs are presented in five groups i.e. up/ down, horizontal, oblique, orbital and lastly miscellaneous actions. In assessing the intensity of an AU the following terminology is used to distinguish different levels of Action Unit involvement. An Action Unit can be totally uninvolved or it can be involved to a maximum level i.e. Trace, Slight, Marked, Pronounced, Severe, Extreme and Maximum.
The personnel/ psychologists involved in FACS describe every observable change in facial movement on the basis of AUs, they do not make emotion judgments rather they simply indicate which AUs moved to produce the changes observed in the face. FACS coders rely on direct observation of movement as well as observed changes in facial morphology such as furrows, creases, bulges of skin etc. to make inferences about which muscles moved. The objective nature of FACS coding makes it a powerful tool which can encode ambiguous and subtle expressions and therefore is potentially more suitable for analyzing the small differences in facial affect.
FACS has become a standard in understanding facial behaviour not only in behavioral science research but in other public sector services such as “screening passengers by observation techniques” (SPOT) programme which is operating at some US airports. Many researchers in developed countries are working on computerized methods for analyzing facial behaviors, including methods for automating FACS coding. Automatic facial coding may eliminate the need for a large coding staff such system which is based on advanced computer science technology automatically tracks faces in a video, extracts geometric and texture features and produces temporal profiles of each facial muscle movement. These profiles are quantified to compute frequencies of single and combined Action Units (AUs) in videos and they can facilitate to forecast the intentions of the subject from its facial expressions.
The ongoing research will prove to be useful in places such as Airports, Railway stations, Banks, Cinemas and other such sensitive places where chances of terrorist activities are high. The automated FACS will alarm the security staff off any ambiguous or malicious personnel movement in the controlled vicinity and hence it will act as thought transference equipment forecasting intentions before time. The law and order situation in Pakistan is not so encouraging, rather we see the situation is deteriorating day by day and terrorist activities are emerged as a common challenge for our country. The government and relevant authorities should enhance the pace of research and development such as discussed in this article so that the society can take benefit from it.
The writer is Electrical Engineer in Karachi Nuclear Power Plant.