The respective role of low and high spatial frequencies in supporting configural and featural processing of faces
The respective role of low and high spatial frequencies in supporting configural and featural processing of faces
复制标题
DOI：
10.1068/p5370
复制
发表时间：
2005-01-01
期刊：
影响因子：
1.7
通讯作者：
Rossion, B
中科院分区：
文献类型：
--
作者：
Goffaux, V;Hault, B;Rossion, B
与其他类别相比，人脸处理的一个显著特征被认为是很大程度上依赖于特征之间的度量关系等构形线索。为了测试低空间频率（lsf）和高空间频率（HSFs）在结构和特征处理中的作用，研究人员向受试者展示了三组人脸，这些人脸经过过滤，以保留低空间频率（低于每面宽8个周期）、高空间频率（高于每面宽32个周期）或整个频谱。他们被要求将两张探针脸中的一张与目标脸相匹配。干扰物探头面与靶面在结构、特征或特征和结构上均不同。当差异在构形水平上时，LSF面优于HSF面。相比之下，由于特征差异，HSF人脸与LSF人脸相比具有很强的性能优势。这些结果支持lsf在人脸的构形处理中起主导作用，而特征处理在很大程度上依赖于hsf。
One distinctive feature of processing faces, as compared to other categories, is thought to be the large dependence on configural cues such as the metric relations among features. To test the role of low spatial frequencies (LSFs) and high spatial frequencies (HSFs) in configural and featural processing, subjects were presented with triplets of faces that were filtered to preserve either LSFs (below 8 cycles per face width), HSFs (above 32 cycles per face width), or the full frequency spectrum. They were asked to match one of two probe faces to a target face. The distractor probe face differed from the target either configurally, featurally, or both featurally and configurally. When the difference was at the configural level, performance was better with LSF faces than with HSF faces. In contrast, with a featural difference, a strong performance advantage was found for HSF faces as compared to LSF faces. These results support the dominant role that LSFs play in the configural processing of faces, whereas featural processing is largely dependent on HSFs.