Abstract
It is essential to distinguish between an imposter and a genuine emotion in certain applications. To facilitate this, the number of features is increased by incorporating physiological signals. Physiological changes in the human body cannot be pretended. Human emotional behavior changes the heart rate, skin resistance, finger temperature, EEG etc. These physiological signal parameters can be measured and included as the final feature vector. The network is to be trained considering all the feature points as inputs with a radial basis activation function at the hidden layer and a linear activation function at the output layer. The two physiological parameters galvanic skin response (GSR) and finger tip temperature (FTT) that are predominant in deciding the emotion of a person are considered in this chapter. The measurements made are transmitted to LabVIEW add-on card for further data processing and analysis. The results obtained are nearer to the reality with a good measure of accuracy.
TopStimulus For Emotions
Emotional content can modify and update the goals and consequently alter the direction of attention to the presented stimuli. Emotions and goals are strongly intertwined in the sense that the immediate relevance of any stimulus to a goal defines the emotionality of the stimulus. The relationship between emotions and the personal goals and concerns of individuals is often suggested to be the basis for emotion elicitation and differentiation by appraisal theorists.
For instance the emotional tag of fear can be attached to a threatening stimulus in so far as the latter can potentially impede the goal of survival. Another example is the emotional tag of happiness that can be assigned to any stimulus that advances the goal of well-being. In a similar fashion numerous emotional tags can be given to stimuli that promote or hinder the attainment of goals ranging from basic individual survival goals to more complex social interaction goals. This vast range of emotions and the related goals is not likely to have been formed concurrently. Rather, emotions evolved from very simple mechanisms that ensured harm avoidance and attainment of vital physical resources into more complex mechanisms that guide complex social behavior. This evolution of emotions may in fact be reflected in the brain systems that generate them, with emotions linked to survival arising from evolutionarily old brain systems. Thus the more primitive emotions would be expected to be elicited by more primitive aspect of the environment, and only at higher levels of evolution would complex classification of stimuli have had related emotions associated with them (Arun Ross & Anil Jain, 2003).
Figure 1
illustrates the system, for measuring Galvanic Skin Response (GSR) and Finger Tip Temperature (FTT). The subject under analysis is made to change the emotional status by external influence. Data are acquired in a closed laboratory condition, by playing the video/audio clippings or showing images. Table 2 shows the different movie clips used to evoke four prominent emotions i.e. happy, surprise, disgust and fear. Sufficient time is allowed in between so that that subject can reach the relaxed state which is considered as the neutral emotional status. Sufficient care is taken so that the subject’s attention is not diverted from the intended task.
Key Terms in this Chapter
Electro Dermal Response: EDRs are changes in the electrical properties of a person’s skin caused by an interaction between environmental events and the individual’s psychological state.
Finger Tip Temperature: Body temperature is one of the known indicators of the general well-being of a person. Two basic types of temperature measurements can be obtained from the human body - systemic and skin surface measurements. Systemic temperature is the temperature of the internal regions of the body. Skin surface temperature is an indicator of the emotional status of a person.
Radial Basis Function: The radial basis function has a maximum of 1 when its input is 0. Used in classification type neural networks. A radial basis neuron acts as a detector that produces 1 whenever the input p is identical to its weight vector w.
Galvanic Skin Response: Human skin is a good conductor of electricity and when a weak electrical current is delivered to the skin, changes in the skin’s conduction of that signal can be measured. The variable that is measured is either skin resistance or its reciprocal, skin conductance which are known as Galvanic Skin Response.
Emotion: Emotion is an excited mental state. Emotion, mood and affective attitude are different but strongly related and influence each other.
LabVIEW: Laboratory Virtual Instrument Engineering Workbench, developed by National Instruments (NI) is a programming environment in which users can create programs using a graphical notation (connecting functional nodes via wires through which data flows).
Electroencephalogram (EEG): The brain generates rhythmical potentials which start off in the individual neurons of the brain. These potentials get added as millions of cell emancipation synchronously and appear as an exterior waveform, the recording of which is known as the electroencephalogram.
Stimulus: Stimulus is the input applied to the subject to induce emotion. It may include playing video or only audio or displaying images in front of a subject.