认知与发展心理学研究室知识库

个体可以从类面孔物体中感知到本不存在的面孔，这种现象被称作面孔视错觉。那些能引起面孔视错觉的物体，被称作类面孔物体。面孔视错觉要素已广泛应用于艺术、广告及商品中，起到吸引注意、促进消费的作用。有关面孔视错觉的研究，有助于揭示大脑是如何构建面部特征从而做出错误判断，为面孔加工机制提供参考。在发生面孔视错觉时，个体可以从类面孔物体中感知到类似面孔的属性，如面部的整体构型或局部特征信息。并且，个体也可以从类面孔物体中感知到社会属性，如面孔情绪或人格特质。此外，类面孔物体可能通过类似眼睛的特征引起注意捕获或导向。然而，目前的研究更多聚焦于类面孔物体的局部特征 在面孔视错觉产生中的作用。基于面孔双通道理论和感官预测模型，本研究认为，在产生面孔视错觉时，个体可能分别加工类面孔物体的整体构型和局部特征，并将记忆和经验中有关面部的属性填充至所输入的视觉信号。再者，类面孔物体引起的注意导向，也可能涉及到上述过程。基于以往有关面孔识别和注意导向的研究，本研究采用了视错觉监测任务和目光线索任务，从感知层面和感知调控层面分别研究了面孔视错觉的发生过程以及类面孔物体所引起的注意导向。

在研究一中，我们使用视错觉监测任务，通过空间频率滤波分离类面孔物体的整体构型（LSF 信号）和局部特征（HSF 信号），并确定两种成分在类面孔感 知中的作用。在实验1中，我们给被试呈现了类面孔物体和匹配物体的 BSF 图片、LSF 图片以及 HSF 图片。我们发现，个体从类面孔物体的 BSF 图片中感知到类似面孔的程度更强，其次是 LSF 图片，最后是 HSF 图片。在实验 2 中，我增加了 LSF & HSF 图片，并采集分析了相关的脑电成分。包括与 LSF 信号更为敏感的 P100、与 HSF 信号更为敏感的 N100，以及涉及认知负荷的 N250。我们发现个体能够从类面孔物体的 LSF 图片和 HSF 图片中产生面孔视错觉，同时相较于 HSF 信号，个体在产生面孔视错觉时更依赖于 LSF 信号。通过脑电成分的分析，我们发现 LSF 图片诱发了更强的 P100，HSF 图片诱发了更强的N100。 同时，类面孔物体的 N250 成分在空间频率上的差异同行为评分一致。P100 和N100成分的结果说明了枕叶皮质对类面孔物体的不同空间频率信号进行初级加工，N250 成分的结果说明了前额叶皮质基于不同空间频率信号所做出的认知资源调用。

在研究二中，我们使用目光线索任务，通过正立呈现（实验 3）、局部保留（实验 4）、倒置呈现（实验 5）和空间频率滤波（实验 6）分理出类面孔物体的两种成分，以此确定两种成分在引起注意导向中的作用。在实验 3 至实验 6，我们均发现，同目光转移面孔类似，类面孔物体作为线索材料也能产生线索效应，效应的产生包括了在一致条件下（线索指向目标出现位置）的注意增益，和在不一致条件下（线索指向目标出现位置对侧）的注意损耗。具体来说，类面孔的整体构型和类似眼睛的局部特征均会引起注意导向。此外，在实验 6 中，我们采集 分析了与HSF 信号更为敏感的 N170 和 P250，以及涉及认知负荷的 N250。通过 脑电成分的分析，我们发现同目光转移面孔类似，类面孔物体也会诱发涉及面孔局部特征加工的 N170 和面孔深层次加工的 P250。同时，在两种线索材料中，均发现目标相关的 N250 成分表现出一致性优势。上述结果进一步说明了枕颞叶皮质对类似眼睛的局部特征进行加工，进而影响到了后续的注意资源调配。

通过一系列实验，我们确认了类面孔物体的整体构型和局部特征在产生面孔视错觉和引起注意导向方面的作用。此外，本研究还为面孔视错觉和面孔识别的机制，以及类面孔元素在商品广告中的应用提供了参考。

Individuals can perceive non-existent faces from face-like objects, a phenomenon known as face pareidolia. Elements of face pareidolia have been widely utilized in art, advertising, and marketing to attract attention and stimulate consumption. Research on face pareidolia contributes to understanding how the brain constructs facial features, thus providing insights into facial processing mechanisms. During face pareidolia, individuals can perceive physical attributes like the global configuration or local features from face-like objects. Additionally, individuals can discern social attributes such as facial emotions or personality traits from these objects. Moreover, face-like objects may capture or orient attention through features resembling eyes. However, current research predominantly focuses on the role of local features of face-like objects in generating face pareidolia. Based on the two-pathways of face processing and sensory prediction models, this study suggests that individuals may separately process the global configuration and local features of face-like objects when making face pareidolia, incorporating memory and experience related to facial attributes into the visual signals. Furthermore, attentional shifts induced by face-like objects may also involve these processes. Building upon previous research on face recognition and attentional shifts, this study employs the pareidolia monitoring paradigm and gaze cueing task to investigate the occurrence of face pareidolia and attentional shifts induced by face-like objects from perceptual and perceptual control perspectives.

In Study 1, we employed the pareidolia monitoring paradigm, separating the global configuration (LSF signal) and local features (HSF signal) of face-like objects through spatial frequency filtering, and determined the roles of these two components in face-like object perception. In Experiment 1, participants viewed BSF, LSF, and HSF images of face-like objects and matched objects. We found that individuals perceived stronger face-like parts from BSF images of face-like objects, followed by LSF images and then HSF images. In Experiment 2, we added LSF & HSF images and analyzed ERP. This included the P100, more sensitive to LSF signals, the N100, more sensitive to HSF signals, and the N250 related to cognitive load. We found that individuals could experience face pareidolia from both LSF and HSF images of face-like objects, with a greater reliance on the LSF signal compared to the HSF signal. ERP analysis revealed stronger P100 responses to LSF images and stronger N100 responses to HSF images. Additionally, the N250 component related to face-like objects showed consistency with behavioral scores. Results of P100 and N100 components indicated that the occipital cortex processes different spatial frequency signals of face-like objects, while N250 component results suggested cognitive resource allocation by the frontal cortex based on different spatial frequency signals.

In Study 2, we employed the gaze cueing task, manipulating the presentation of face-like objects through upright presentation (Experiment 3), local retention (Experiment 4), inverted presentation (Experiment 5), and spatial frequency filtering (Experiment 6), to determine the roles of these two components in inducing attentional shifts. Across Experiments 3 to 6, we consistently observed cueing effects similar to those of averted gaze faces, suggesting that face-like objects also elicited cueing effects as cueing materials. These effects included attentional benefits under congruent trials (cue pointing to the target appearing position) and attentional costs under incongruent trials (cue pointing to the opposite side of the target appearing position). Specifically, both the global configuration and eye-like local features of objects induced attentional shifts. Furthermore, in Experiment 6, we collected and analyzed N170 and P250 components more sensitive to HSF signals, and the N250 component related to cognitive load. EEG component analysis revealed that similar to averted gaze faces, face-like objects also elicited N170 related to facial local feature processing and P250 related to deeper facial processing. Additionally, in both cueing materials, target-related N250 components exhibited consistency advantages. These results further indicated that the occipitotemporal cortex processes eye-like local features, influencing subsequent attentional resource allocation.

Through a series of experiments, we confirmed the roles of the global configuration and local features of face-like objects in inducing face pareidolia and attentional shifts. Additionally, this study provides insights into the mechanisms of face pareidolia and face recognition, as well as the application of face-like elements in advertising.