Zhang Xukun, Sun Yusheng, Liu Wei, Zhang Zhijun, Wu Binxing
Department of Psychology and Behavioral Sciences, Zhejiang University, Hangzhou, China.
Department of Psychology and Behavioral Sciences, Zhejiang University, Hangzhou, China; School of Education, Yunnan Minzu University, Kunming, Yunnan, China.
Acta Psychol (Amst). 2020 Jul;208:103101. doi: 10.1016/j.actpsy.2020.103101. Epub 2020 May 30.
The low spatial frequency (LSF) component of visual information rapidly conveyed coarse information for global perception, while the high spatial frequency (HSF) component delivered fine-grained information for detailed analyses. The feedforward theorists deemed that a coarse-to-fine process was sufficient for a rapid scene recognition. Based on the response priming paradigm, the present study aimed to deeply explore how different spatial frequency interacted with each other during rapid scene recognition. The response priming paradigm posited that as long as the prime slide could be rapidly recognized, the prime-target system was behaviorally equivalent to a feedforward system. Adopting broad spatial frequency images, experiment 1 revealed a typical response priming effect. But in experiment 2, when the HSF and the LSF components of the same pictures were separately presented, neither the LSF-to-HSF sequence nor the HSF-to-LSF sequence reproduced the response priming effect. These results demonstrated that LSF or HSF component alone was not sufficient for rapid scene recognition and, further, that the integration of different spatial frequency needed some early feedback loops. These findings supported that the local recurrent processing loops among early visual cortex was involved during rapid scene recognition.
视觉信息的低空间频率（LSF）成分迅速传递粗略信息以进行整体感知，而高空间频率（HSF）成分则传递细粒度信息以进行详细分析。前馈理论学家认为，从粗略到精细的过程足以实现快速场景识别。基于反应启动范式，本研究旨在深入探讨在快速场景识别过程中不同空间频率是如何相互作用的。反应启动范式假定，只要启动幻灯片能够被快速识别，启动-目标系统在行为上就等同于前馈系统。实验1采用宽空间频率图像，揭示了典型的反应启动效应。但在实验2中，当同一图片的HSF和LSF成分分别呈现时，无论是从LSF到HSF的顺序还是从HSF到LSF的顺序，都没有重现反应启动效应。这些结果表明，单独的LSF或HSF成分不足以实现快速场景识别，进一步表明不同空间频率的整合需要一些早期反馈回路。这些发现支持了早期视觉皮层之间的局部循环处理回路在快速场景识别过程中发挥了作用。
