人类在检测和辨别生物运动方面的效率。

Human efficiency in detecting and discriminating biological motion.

作者信息

Lu Hongjing, Tjan Bosco S, Liu Zili

机构信息

Department of Psychology, University of California, Los Angeles, Los Angeles, CA, USADepartment of Statistics, University of California, Los Angeles, Los Angeles, CA, USA.

Department of Psychology and Neuroscience Graduate Program, University of Southern California, Los Angeles, CA, USA.

出版信息

J Vis. 2017 Jun 1;17(6):4. doi: 10.1167/17.6.4.

DOI:10.1167/17.6.4

PMID:28593248

Abstract

Using an "information meter" provided by ideal observer analysis, we measured the efficiency with which human observers processed different walking stimuli against luminance noise and spatial uncertainty to either detect the presence of a walker or to discriminate the walking direction. Human efficiency was examined across four renderings of a human walker: contour, point lights, silhouette, and skeleton. We replicated the previous finding of low discrimination efficiency in biological motion (Gold, Tadin, Cook, & Blake, 2008) and also found low detection efficiency for biological motion. Interestingly, in both detection and discrimination tasks, the skeleton display was among those yielding the highest level of efficiency in processing visual information. This finding suggests that structural information about the relative position of joints, highlighted in the skeleton display, provides a critical component of the internal representation for biological motion.

摘要

利用理想观察者分析提供的“信息测量仪”，我们测量了人类观察者在存在亮度噪声和空间不确定性的情况下处理不同行走刺激以检测步行者的存在或辨别行走方向的效率。我们对人类步行者的四种渲染形式进行了人类效率测试：轮廓、点光源、剪影和骨架。我们重复了之前关于生物运动中低辨别效率的研究结果（戈尔德、塔丁、库克和布莱克，2008年），并且还发现了生物运动的低检测效率。有趣的是，在检测和辨别任务中，骨架显示都是处理视觉信息效率最高的形式之一。这一发现表明，骨架显示中突出的关节相对位置的结构信息为生物运动的内部表征提供了关键组成部分。