Budget Amount *help |
¥3,600,000 (Direct Cost: ¥3,600,000)
Fiscal Year 1999: ¥500,000 (Direct Cost: ¥500,000)
Fiscal Year 1998: ¥3,100,000 (Direct Cost: ¥3,100,000)
|
Research Abstract |
Faces are hard to recognize if they are shown upside down or in photographic negatives. This suggests that the face recognition process might be somewhat special. The present study mainly aimed to investigate the specificity of information processing in the early stages of face recognition. The study consisted of three parts, focusing on two-dimensional image processing, three-dimensional image processing and the effect of image properties on the higher-order cognitive process. At first, two-dimensional aspects of face recognition were considered through a review of previous studies on the spatial characteristic of face recognition and an experiment on the effect of image attributes on human face detection. As the result, the phenomenal peculiarity of face recognition is considers due to the characteristics of information acquisition process, and such peculiarity is observed even in the face detection task, that requires more primitive manipulation than usually used face identification
… More
tasks. In the second part, three-dimensional aspects of face recognition process were investigated through two experiments : one was concerned with the 'hollow-face' illusion and the other with the 'mental rotation' of faces. The results from these experiments showed that two-dimensional facial textures were crucial for recognition and they affected perceived structure of the surfaces. These implied that the attributes of facial images affect the lower-order perceptual process. In the third part, the effect of facial expression included in the images on the higher-order cognitive process, such as the evaluation of intention in conversation, was examined. Although this part was not the main stream of the present study, results showed facial expression affected the evaluation significantly, implying visually mediated facial attributes affect the high-order cognitive process. Through these studies, it is suggested that the current face recognition models should be detailed extensively, covering lower- and higher-order processes of visual cognition. Less
|