Facial Expression Information Research Articles

ニューラルネットによる人の基本表情認識

An attempt is being made to develop a mind-implemented robot that can carry out intellectual conversation with humans. As the first step for this study, a method for the robot to perceive human emotion is investigated. Specifically, a method of human emotion recognition from facial expressions by a neural network is shown. The authors categorized facial expressions into six groups (surprise, fear, disgust, anger, happiness, and sadness) and obtained 70 CCD (charge coupled device) camera-acquired data of facial feature-points from three components of the face (eyebrows, eyes, and mouth). Then the facial expression information is generated and input into the input units of the neural network; network learning and recognition are done by a backpropagation algorithm. >

A Nose Gesture Interface Device: Extending Virtual Realities

This paper reports on the1 development of a nose-machine interface device that provides real-time gesture, position, smell and facial expression information. The DATANOSE™2—Data AtomaTa CORNUCOPIA pNeumatic Olfactory 1/O-deviSE Tactile Manipulation (Olsen, 1986; Myers, 1991)—allows novice users without any formal nose training to perform complex interactive tasks.

A nose gesture interface device

This paper reports on the1 development of a nose-machine interface device that provides real-time gesture, position, smell and facial expression information. The DATANOSETM2-Data AtomaTa CORNUCOPIA...

In and out of context: Influences of facial expression and context information on emotion attributions

Studies intended to evaluate the relative importance of person information (facial expression) and context information (situation evoking an emotion) in determining emotion attributions of observers have so far, with a few exceptions, used static stimulus presentation (photographs and/or verbal descriptions). Here an attempt is made to study emotion inference processes by using dynamic presentation of stimuli (video). Sixty judges were confronted with 60 video clips depicting an emotional situation followed by an emotional facial expression of an actor/actress in reaction to this situation. These clips were selected from films and TV shows. Three groups of 20 judges each watched (i) only the first takes of the clips depicting an emotion‐arousing situation, or (ii) only the second takes of the clips presenting the emotional facial expressions, or (iii) combinations of both takes depicting the emotion‐arousing situation followed by the facial expression. Their task was to judge the emotion(s) expressed by the person in the given situation. Results indicate that context information dominated person information in determining emotion attributions. Furthermore, differences were found with respect to the relative discrepancy or consonance of clips (situation information was more important to judges in discrepant compared with consonant clips) and differences due to gender of the actors (when watching actresses, person information dominated judgements; when watching actors, situation information was dominant). Results are discussed with respect to possible differences between static and dynamic stimulus presentation and with respect to the general process of inferring emotions.

The regulation of infant behavior by maternal facial expression

In the past, Infants' comprehension of adult facial expressions has been assessed through discrimination and categorization techniques. However, a more meaningful level of comprehension involves infants' ability to utilize the information communicated by facial expressions to regulate their instrumental behavior. This study explored the age at which the capacity for such utilization develops. Seventy-two infants, half 12 months and half 18 months, were each presented with three unusual toys designed to elicit uncertainty in the infants. When they visually referenced their mothers' faces, mothers posed smiling, fearful, or neutral expressions. At both ages, the infants moved closest to the mother when she posed fear, moved farthest from her when she posed joy, and maintained an intermediate distance when she appeared neutral. For all variables, the effect was most clear on the third trial. The results demonstrated that by 12 months of age infants seek out facial expression information and use it to regulate their own behavior.