Sperling George, Sun Peng, Liu Dantian, Lin Ling
Department of Cognitive Sciences, University of California, Irvine.
Psychol Rev. 2020 Apr;127(3):305-326. doi: 10.1037/rev0000180.
At an early stage, 3 different systems independently extract visual motion information from visual inputs. At later stages, these systems combine their outputs. Here, we consider a much studied (>650 publications) class of visual stimuli, plaids, which are combinations of 2 sine waves. Currently, there is no quantitative theory that can account for the perceived motion of plaids. We consider only perceived plaid direction, not speed, and obtain a large set of data exploring the various dimensions in which same-spatial-frequency plaids differ. We find that only 2 of the 3 motion systems are active in plaid processing, and that plaids with temporal frequencies 10 Hz or greater typically stimulate only the first-order motion system, which combines the plaid components by vector summation: Each plaid component is represented by a contrast-strength vector whose length is contrast-squared times a factor representing the relative effectiveness of that component's temporal frequency. The third-order system, which becomes primary at low temporal frequencies, also represents a plaid as 2 vectors that sum according to their contrast strength: a pure plaid in which both components have equal contrast and a residual sine wave. Second-order motion is irrelevant for these plaids. These principles enable a contrast-strength-vector summation theory for the responses of the first-order and third-order motion systems. With zero parameters estimated from the data, the theory captures the essence of the full range of the plaid data and supports the counterintuitive hypothesis that motion direction is processed independently of speed at early stages of visual processing. (PsycInfo Database Record (c) 2020 APA, all rights reserved).
在早期阶段,3种不同的系统独立地从视觉输入中提取视觉运动信息。在后期阶段,这些系统会整合它们的输出。在此,我们考虑一类经过大量研究(>650篇出版物)的视觉刺激,即格子图案,它是由2个正弦波组合而成。目前,尚无能够解释格子图案感知运动的定量理论。我们仅考虑感知到的格子图案方向,而非速度,并获得了一大组数据,探索相同空间频率的格子图案在不同维度上的差异。我们发现,在格子图案处理过程中,3种运动系统中只有2种是活跃的,并且时间频率为10赫兹或更高的格子图案通常仅刺激一阶运动系统,该系统通过矢量求和来组合格子图案的成分:每个格子图案成分由一个对比度强度矢量表示,其长度是对比度的平方乘以一个代表该成分时间频率相对有效性的因子。三阶系统在低时间频率时起主要作用,它也将一个格子图案表示为2个根据其对比度强度求和的矢量:一种是两个成分具有相等对比度的纯格子图案和一个残余正弦波。二阶运动与这些格子图案无关。这些原理促成了一种关于一阶和三阶运动系统响应的对比度强度矢量求和理论。该理论在未根据数据估计任何参数的情况下,捕捉到了整个格子图案数据范围的本质,并支持了一个违反直觉的假设,即在视觉处理的早期阶段,运动方向是独立于速度进行处理的。(PsycInfo数据库记录(c)2020美国心理学会,保留所有权利)