Understanding Facial Expressions - Case Study Example

Understanding Facial Expressions Introduction Facial expressions are distinct human features which set us apart from the other creatures in the animal kingdom. For every emotion we feel – from the happy to the sad to the pained and to the ecstatic, there are various facial expressions that we can manifest. This paper shall discuss facial expressions and how we can understand and interpret them from a psychological point of view. Various psychology journals shall be used in order to explain and evaluate these facial expressions. Discussion In a study by Calder and Young (2005), they sought to understand the recognition of facial identity and facial expressions. They first pointed out that our face can portray different emotions. Previous researches have established that the recognition process of facial expressions involves “separable visual pathways imaging and cell-recording studies [which] are commonly interpreted within this framework” (Calder and Young, 2005). In other words, facial expressions are interpreted based on imaging and on visual measure patterns. There are however, other pathways or manners of interpretation which may be used in order to understand facial expressions. The authors make mention of the fact that there are other measures or tools which are now available and which are more accurate in evaluating facial expressions. The authors also pointed out that expressions are processed through separate pathways – from what we get to see to the mechanisms that we use for their interpretation. However, this manner of interpretation is not adequately supported by evidence. They have pointed out the importance of an approach to face perception which would emphasize the different physical properties of the different facial characteristics. This approach is distinct from the classic approach which tends to emphasize differentiation mainly on informational content (Calder and Young, 2005). In a discussion by Ekman, (as quoted by Dalgeish and Power, 1999) he pointed out that universal expressions of emotion “is the connection between particular facial configurations and specific emotions”. Although, universal expressions of emotion relate to specific facets of emotions, expressions may not always manifest because most people are capable of controlling their expressions. In another way, emotions may not always occur when we manifest facial expressions because most people are capable of fabricating expressions. Ekman’s discussion goes on by saying that the connection between emotions and expressions was developed through natural selection and through constant learning. It is therefore difficult to match particular expressions with specific emotions because similar expressions may be used to express different emotions. Specific variations may be seen in instances of extreme opposite reactions, such as anger, disgust, sadness, and fear. However, it is important to factor in the culture from which these expressions are seen. For example, contempt is a universal emotion, however, in some cultures, it is often not encouraged as an expressed emotion (Ekman, as quoted by Dalgeish and Power, 1999). And even in saying that there is a connection between facial expressions and emotions, such is still an incomplete statement because it still does not specify what sort of emotion is actually connected. In one instance, it may actually be the message which is interpreted when gazing into the face of the other person or it may actually be the feelings, the biological changes, the memories that the person is experiencing which is creating the expression. As was mentioned, culture, social groupings and individual differences tend to create various differences in the facial expression of emotions (Ekman, as quoted by Dalgeish and Power, 1999). The words used by a person and the words used to portray such emotions are very much different from each other, and these differences affect our emotional experiences as human beings. Zhang and Ji (2003) explored the use of multisensory information fusion technique with Dynamic Bayesian Networks (DBNs) in order to model and understand the temporal behaviors of imaging sequences in facial expressions. The authors attempted to understand facial expressions by using a probabilistic framework by combining the DBNs with the facial action units (FAUs) from a psychological point of view. The DBNs give access to logical and organized probabilistic framework which represents spatial and temporal data related to facial expressions and to make possible a selection of the most informative visual elements from the information gathered (Zhang and Ji, 2003). From their research, Zhang and Ji (2003) were able to establish that recognizing and understanding facial expressions may be done by integrating current visual observations with previous visual evidences. As a result, it is now possible to recognize and understand more accurately the temporal behavior of facial expressions. Through experimental results, it is possible to demonstrate the accuracy and the admissibility of Zhang and Ji’s approach for facial expression analysis in image sequences (2003). Zhang and Ji (2003) are quick to emphasize that the current available tool using spatio-temporal analysis in understanding facial expressions is fraught with various flaws. Computing the dense flow between successive images is an ineffective manner of understanding facial expressions because flow estimates can be easily disturbed by lighting, by non-rigid motion, and by inaccurate image registration (Zhang and Ji, 2003). The facial flow estimates takes into consideration the beginning, the peak, and the end of the expression, however, it does not measure the facial movement itself. Consequently, the measure of facial expression does not reflect the evolution and the possible momentary intensity of an observed expression (Zhang and Ji, 2003). The methods proposed by Zhang and Ji (2003) focus on capturing correlations and dependencies in facial expressions in order to allow for dynamic models of facial expressions. The recognition system is based on actively selecting the details of the most important facial visual cues at particular points in time and correlating these with previous visual cues (Zhang and Ji, 2003). Consequently, Zhang and Ji (2003) have taken advantage of DBNs and other sensory information in order to allow a well-rounded analysis of facial expressions. In a study by Smith, Cottrell, Gosselin, and Schyns (2005), they sought to assess the human face as a transmitter of expression signals and the brain as a decoder of these signals. They set forth that when the face evolves to take full advantage of the transmission of these expression signals, then the primary facial expressions should then have a basic overlap in their transmission. And if the brain has also evolved to optimize the classification of expressions, then it should be efficient in the interpretation of information from the transmitter (Smith, Cottrell, Gosselin, and Schyns, 2005). In their study, they described the data related to the identification of the six basic facial expression signals and to assess how each expression is broken down by the brain. After comparing the diagnostic filtering functions of the human and the model observers, the researchers set forth that the face has indeed evolved to transmit expression signals that have minimal relationships with each other; the brain then further sets apart the correlation and improves the transmission and interpretation of these signals. These signals make up maximum inputs which can then be used to set apart specific response of specialized brain structures (Smith, Cottrell, Gosselin, and Schyns, 2005). The researchers emphasize that the emotional expressions that the face can transmit is actually often limited by the skeletal and muscular movements of the face as an encoder. It is also constrained by “pressures of decoding expressions from long viewing distances, and by generic computational requirement of transmitting decorrelated signals” (Smith, Cottrell, Gosselin, and Schyns, 2005). In a 1995 study, Essa and Pentland attempted to develop new and more accurate “representations for facial expressions by building a video database of facial expressions and then probabilistically characterizing the facial muscle activation associated with each expression using a detailed physical model of the skin and muscles”. They did their study because they noticed that the prior techniques at facial expression recognition were often based on the Facial Action Coding System which is a technique developed to allow psychologists to code expressions from static facial pictures or ‘mugshots’ (Essa and Pentland, 1995). Their proposed method is capable of producing a muscle-based representation of facial movements which will then be assessed to recognize facial expressions in two different ways. The researchers established that the first method would use the physics-based model by evaluating expressions and by comparing estimated muscle activations; the second method would use the physics-based model to generate spatio-temporal motion energy templates of the entire face and of each expression (Essa and Pentland, 1995). Through these simple and biologically-possible templates which are ‘powered’ by energy, facial recognition of expressions is possible and even more accurate. These methods proposed and used by these researchers in the course of their study indicated substantially better accuracy at recognizing expressions as compared to previous attempts by previous methods and techniques (Essa and Pentland, 1995). This study has presented the groundwork in the 1990s for more accurate models of recognizing and understanding facial expressions. This method will actually be followed up by other more accurate models in recognizing facial expressions. Nevertheless, this study was able to point out that the original models do not actually measure the particular nuances of facial expressions. Finally, in a study by Jack, Blais, Scheepers, Schyns, and Caldara (2009) they sought out to establish that cultural confusions indicate that facial expressions are not actually universal. They first acknowledged in their study that the interaction of human beings on a global scale is possible through biologically rooted social signs – facial expressions of emotions. However, they also point out in the paper that some negative facial expressions are sometimes not recognized among Eastern groups. In their study, they focused on decoding or assessing facial expression signals by merging behavioral and computational analyses with new spatiotemporal analyses of eye movements (Jack, Blais, Scheepers, Schyns, and Caldara, 2009). Through their analyses, they were able to establish that Eastern observers use culture-specific decoding method which inadequately distinguishes universal facial expressions like fear and disgust. The researchers further extrapolated that Easterners usually fixated their expressions on the eye region, in contrast to the Westerners who usually distributed their fixations evenly across their face (Jack, Blais, Scheepers, Schyns, and Caldara, 2009). Consequently, their study revealed that as Easterners concentrated on their eyes, the researchers often came up with inaccurate and ambiguous information. These researchers then pointed out through their study results that a universal interpretation and model of understanding facial expressions cannot be truly applied because facial expressions are complex and are influenced by culture and by globalization (Jack, Blais, Scheepers, Schyns, and Caldara, 2009). The studies above emphasize that there are different models which may be used in order to measure and understand facial expressions. These models have evolved through different researches and through the better understanding that we now have of human expression and communication. More importantly, in understanding facial expression, attempts should be made in the context of specificity and recognition of the impact of culture. Works Cited Calder, A. and Young, A. (2005) Understanding the recognition of facial identity and facial expression. Sburover. Retrieved 09 November 2009 from http://www.sburover.it/psice/psicologia/valutazione_cognitiva/Riconoscimento_volti.pdf Dagleish, T and Power, M. (1999) Handbook of Cognition and Emotion. New York: John Wiley and Sons Ekman, P. (n.d) Chapter 16: Facial Expressions. Virtual Human Markup Language. Retrieved 09 November 2009 from http://www.vhml.org/theses/wijayat/sources/writings/papers/facial_expression.pdf Essa, I. and Pentland, A. (1995) Facial Expression Recognition using a Dynamic Model and Motion Energy. Computer Vision Proceedings International Conference Jack, R., Blais, C., Scheepers, C., Schyns, P., and Caldara, R. (2009) Cultural Confusions Show that Facial Expressions Are Not Universal. Current Biology, volume 19, issue 18, pp. 1543-1548 Smith, M., Cottrell, G., Gosselin, F., Schyns, P. (2005) Research Report: Transmitting and Decoding Facial Expressions. University of Montreal. Retrieved 09 November 2009 from http://mapageweb.umontreal.ca/gosselif/labogo/emotions_pub.pdf Zhang, Y. and Ji, Q. (2003) Facial Expression Understanding in Image Sequences Using Dynamic and Active Visual Information Fusion. International Conference on Computer Vision. Retrieved 09 November 2009 from http://www.ecse.rpi.edu/~qji/Papers/1297_zhang.pdf Read More