空间频率与视野位置在面孔识别中的作用

发布时间：2018-06-04 15:27

本文选题：空间频率 + 视野位置　；参考：《河北大学》2014年硕士论文

【摘要】：面孔在人们的生活中发挥着重要的社会功能，个体通过对面孔刺激的加工可以提取到各种社会信息，如年龄、性别、种族、表情等。White将面孔信息区分为构型信息与特征信息来解释身份表情信息之间的分离：身份识别依靠的多是构型信息，而表情识别更多的依靠特征信息。但是这样的划分存在局限性。目前支持身份与表情双分离的尺度并不多。以往的研究显示，高频空间信息倾向于对细节部分的表征，如表情纹、鼻子的轮廓、眼睛、嘴巴。低频空间信息不具备单独的表征某个器官的分辨率，但是可以为面孔的探测提供信息。因为对高频信息敏感的视锥细胞主要集中在视网膜中央窝，对低频信息敏感的视杆细胞主要集中在视网膜外周，因此空间频率和视野位置在面孔识别中可能存在着交互作用，空间频率可能是支持面孔表情加工与身份加工是分离的指标之一。本研究通过事件相关电位技术，选取N170作为主要的观测指标，分别从面孔的表情识别和身份识别上考察了空间频率和视野位置对面孔识别的作用。实验被试为来自河北大学的20名本科生及研究生。实验一要求被试对随机出现在视野中央或外周的高低频面孔表情进行判断，高兴按鼠标左键，愤怒按鼠标右键；实验二中要求被试判断面孔的身份，左键代表名人，右键代表非名人。实验过程中记录被试的行为数据和脑电数据。实验结果显示： 1．在实验一中，空间频率与视野位置存在交互作用。高频面孔在视野中央诱发的N170潜伏期比低频面孔短，波幅小于低频面孔；在视野外周时潜伏期长，波幅与低频面孔差异不显著。 2．在实验二中，，空间频率与视野位置交互作用不显著。高频面孔在视野中央诱发的N170潜伏期比低频面孔长，波幅大于低频面孔。高频面孔在视野外周时潜伏期较短，波幅差异不显著。 3．高频信息对面孔的表情识别作用优于低频信息；面孔的身份识别主要依赖于面孔图片的低频信息。
[Abstract]:Faces play an important social role in people's lives. Individuals can extract a variety of social information, such as age, sex, race, through the processing of face stimulation. White divides face information into configuration information and feature information to explain the separation of identity expression information: identity recognition depends on configuration information and facial expression recognition depends more on feature information. However, such a division has limitations. At present, there are few scales to support the separation of identity and expression. Previous studies have shown that high-frequency spatial information tends to represent the details, such as facial lines, nose contours, eyes and mouth. Low frequency spatial information does not have the resolution to represent an organ alone, but it can provide information for face detection. Because the cone cells sensitive to high frequency information are mainly concentrated in the central retinal fossa and the rod cells sensitive to low frequency information are mainly concentrated on the periphery of the retina, there may be interaction between spatial frequency and visual field position in face recognition. Spatial frequency may be one of the indicators supporting the separation of facial expression processing and identity processing. The effects of spatial frequency and visual field position on face recognition were investigated from facial expression recognition and identity recognition. The subjects were 20 undergraduates and postgraduates from Hebei University. In experiment one, participants were asked to judge the high-low frequency facial expressions that appeared randomly in the center of the field of vision or around them. They were happy to press the left mouse button and angrily to press the right mouse button. In experiment two, the subjects were asked to judge the identity of the face, and the left button represented the celebrity. Right-right represents non-celebrity. The behavioral data and EEG data of the subjects were recorded during the experiment. In experiment 1, there is interaction between spatial frequency and visual field position. The latency of N170 induced by high frequency faces in the center of visual field is shorter than that of low frequency faces, and the amplitude of N170 is smaller than that of low frequency faces. In experiment 2, the interaction between spatial frequency and visual field position was not significant. The latency of N170 induced by high frequency faces in the center of field is longer than that of low frequency faces, and the amplitude is larger than that of low frequency faces. The latency of high frequency faces was shorter and the amplitude difference was not significant. 3. 3. The high frequency information is superior to the low frequency information in facial expression recognition, and the face identification mainly depends on the low frequency information of the face image.
【学位授予单位】：河北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：B842

【参考文献】