Two routes to face perception: evidence from psychophysics and computational modeling

The aim of this study was to separately analyze the role of featural and configural face representations. Stimuli containing only featural information were created by cutting the faces into their parts and scrambling them. Stimuli only containing configural information were created by blurring the faces. Employing an old-new recognition task, the aim of Experiments 1 and 2 was to investigate whether unfamiliar faces (Exp. 1) or familiar faces (Exp. 2) can be recognized if only featural or configural information is provided. Both scrambled and blurred faces could be recognized above chance level. A further aim of Experiments 1 and 2 was to investigate whether our method of creating configural and featural stimuli is valid. Pre-activation of one form of representation did not facilitate recognition of the other, neither for unfamiliar faces (Exp. 1) nor for familiar faces (Exp. 2). This indicates a high internal validity of our method for creating configural and featural face stimuli. Experiment 3 examined whether features placed in their correct categorical relational position but with distorted metrical distances facilitated recognition of unfamiliar faces. These faces were recognized no better than the scrambled faces in Experiment 1, providing further evidence that facial features are stored independently of configural information. From these results we conclude that both featural and configural information are important to recognize a face and argue for a dual-mode hypothesis of face processing. Using the psychophysical results as motivation, we propose a computational framework that implements featural and configural processing routes using an appearance-based representation based on local features and their spatial relations. In three computational experiments (Experiments 4–6) using the same sets of stimuli, we show how this framework is able to model the psychophysical data.     

面孔知觉中特征、结构和整体加工策略的眼动研究

考察面孔知觉中特征加工、结构加工和整体加工三种不同加工方式对应的眼动模式。实验1中, 将以特征信息为主的错乱面孔和以结构信息为主的模糊面孔作为线索刺激, 引发对测试面孔的特征加工和结构加工, 眼动分析表明：特征加工表现为对面孔各特征内更长的凝视时间, 结构加工表现为对面孔各特征间高频的眼跳。实验2采用相同的研究范式, 将完整面孔、轻度错乱面孔和低水平模糊面孔作为线索刺激, 引发对测试面孔除特征加工和结构加工外的另一种加工方式-- 整体加工, 表现为注视点更多地落在测试面孔中央区的鼻子部位以扩大注视范围, 进而把握整张面孔信息。本研究揭示了三种不同面孔加工方式眼动模式的差异。     

眼睛区域构型信息与特征信息的跨维共变增益效应及其加工特异性

面孔知觉可能在区域尺度上发生多维信息整合, 但迄今无特异性实验证据。本研究在两个实验中操纵面孔眼睛区域或嘴巴区域的单维构型或特征信息, 测量人们觉察单维变化或跨维共变的敏感度, 以此检测面孔区域尺度上的多维信息整合有何现象与规律, 进而揭示面孔知觉的多维信息整合机制。实验获得3个发现：(1)正立面孔眼睛区域的信息变化觉察呈现出“跨维共变增益效应”, 跨维信息共变觉察的敏感度显著高于任意一种单维信息变化觉察的敏感度; (2)“跨维共变增益效应”只在正立面孔的眼睛区域出现, 在倒置面孔的眼睛区域、正立面孔的嘴巴区域或倒置面孔的嘴巴区域都没有出现, 因此具有面孔区域特异性和面孔朝向特异性; (3)就单维信息变化觉察而言, 眼睛区域的敏感度不会受到面孔倒置的损伤, 但是嘴巴区域的敏感度会受到面孔倒置的显著损伤。综合可知, 面孔知觉确实会发生区域尺度上的信息整合, 它不是普遍性的信息量效应, 而是特异性的眼睛效应(只发生在正立面孔的眼睛区域)。这是面孔整体加工(face holistic processing)在单维信息分辨和多维信息整合之间建立联系的必经环节。这提示我们对全脸多维信息知觉整合的理解需要从传统的面孔整体加工假设升级到以眼睛为中心的层级化多维信息整合算法(a hierarchical algorithm for multi-dimensional information integration)。     

Understanding the role of configural processing in face emotion recognition in Parkinson’s disease

This investigation examined whether impairment in configural processing could explain deficits in face emotion recognition in people with Parkinson’s disease (PD). Stimuli from the Radboud Faces Database were used to compare recognition of four negative emotion expressions by older adults with PD (n = 16) and matched controls (n = 17). Participants were tasked with categorizing emotional expressions from upright and inverted whole faces and facial composites; it is difficult to derive configural information from these two types of stimuli so featural processing should play a larger than usual role in accurate recognition of emotional expressions. We found that the PD group were impaired relative to controls in recognizing anger, disgust and fearful expressions in upright faces. Then, consistent with a configural processing deficit, participants with PD showed no composite effect when attempting to identify facial expressions of anger, disgust and fear. A face inversion effect, however, was observed in the performance of all participants in both the whole faces and facial composites tasks. These findings can be explained in terms of a configural processing deficit if it is assumed that the disruption caused by facial composites was specific to configural processing, whereas inversion reduced performance by making it difficult to derive both featural and configural information from faces.     

Thatcher's children: development and the Thatcher illusion

Children do not show the same recognition disadvantage for inverted faces as adults do. It has been suggested that this is because the configural encoding (which is disrupted by inversion) becomes more useful or available as we get older. The distinction between configural processing and featural processing, however, is not always clear--it may be a dichotomy or a continuum. The perceived normal-to-grotesque switch in the Thatcher illusion was investigated, as the image was rotated, by people aged between 6 and 75 years. No effect of age was found, with young children showing the same effects as adults--the switch occurring at about 72 degrees. The development of face processing and the nature of facial configural encoding are discussed in the light of this result.     

The other-race effect: Holistic coding differences and beyond

We evaluate claims that the other-race effect in face memory reflects stronger holistic coding of own-race than other-race faces. Considering evidence from a range of paradigms, including the inversion effect, part–whole effect, composite effect, and the scrambled/blurred task, we find considerable inconsistency, both between paradigms and between participant ethnicities. At the same time, however, studies that isolate configural and component feature processing consistently show better featural, as well as better configural, processing of own-race faces, for both Caucasian and Asian participants. These results raise the possibility that the key feature of own-race face processing is not stronger holistic processing per se, but rather more effective processing of all types of face information (featural as well as holistic).     

Effects of geometric distortions on face-recognition performance

The importance of 'configural' processing for face recognition is now well established, but it remains unclear precisely what it entails. Through four experiments we attempted to clarify the nature of configural processing by investigating the effects of various affine transformations on the recognition of familiar faces. Experiment 1 showed that recognition was markedly impaired by inversion of faces, somewhat impaired by shearing or horizontally stretching them, but unaffected by vertical stretching of faces to twice their normal height. In experiment 2 we investigated vertical and horizontal stretching in more detail, and found no effects of either transformation. Two further experiments were performed to determine whether participants were recognising stretched faces by using configural information. Experiment 3 showed that nonglobal vertical stretching of faces (stretching either the top or the bottom half while leaving the remainder undistorted) impaired recognition, implying that configural information from the stretched part of the face was influencing the process of recognition--ie that configural processing involves global facial properties. In experiment 4 we examined the effects of Gaussian blurring on recognition of undistorted and vertically stretched faces. Faces remained recognisable even when they were both stretched and blurred, implying that participants were basing their judgments on configural information from these stimuli, rather than resorting to some strategy based on local featural details. The tolerance of spatial distortions in human face recognition suggests that the configural information used as a basis for face recognition is unlikely to involve information about the absolute position of facial features relative to each other, at least not in any simple way.     

Features and their configuration in face recognition

Tanaka and Farah (1993) have proposed a holistic approach to face recognition in which information about the features of a face and their configuration are combined together in the face representation. An implication of the holistic hypothesis is that alterations in facial configuration should interfere with retrieval of features. In four experiments, the effect of configuration on feature recognition was investigated by creating two configurations of a face, one with eyes close together and one with eyes far apart. After subjects studied faces presented in one of the two configurations (eyes-close or eyes-far), they were tested for their recognition of features shown in isolation, in a new face configuration, and in the old face configuration. It was found that subjects recognized features best when presented in the old configuration, next best in the new configuration, and poorest in isolation. Moreover, subjects were not sensitive to configural information in inverted faces (Experiment 2) or nonface stimuli (i.e., houses; Experiments 3 and 4). Importantly, for normal faces, altering the spatial location of the eyes not only impaired subjects’ recognition of the eye features but also impaired their recognition of the nose and mouth features—features whose spatial locations were not directly altered. These findings emphasize the interdependency of featural and configural information in a holistic face representation.     

The first 100 milliseconds of a face: on the microgenesis of early face processing

Face recognition involves both processing of information relating to features (e.g., eyes, nose, mouth, hair, i.e., featural processing), as well as the spatial relation between these features (configural processing). In a sequential matching task, participants had to decide whether two faces that differed in either featural or relational aspects were identical or different. In order to test for the microgenesis of face recognition (the development of processing onsets), presentation times of the backward-masked target face were varied (32, 42, 53, 63, 74, 84, or 94 msec.). To test for specific processing onsets and the processing of different facial areas, both featurally and relationally modified faces were manipulated in terms of changes to one facial area (eyes or nose or mouth), two, or three facial areas. For featural processing, an early onset for the eyes and mouth was at 32 msec. of presentation time, but a late onset for the nose was detected. For relationally differing faces, all onsets were delayed.     

Configural face encoding and spatial frequency information

Configural relations and a critical band of spatial frequencies (SFs) in the middle range are particularly important for face recognition. We report the results of four experiments in which the relationship between these two types of information was examined. In Experiments 1, 2A, and 2B, the face inversion effect (FIE) was used to probe configural face encoding. Recognition of upright and inverted faces and nonface objects was measured in four conditions: a no-filter condition and three SF conditions (low, medium, and high frequency). We found significant FIEs of comparable magnitudes for all frequency conditions. In Experiment 3, discrimination of faces on the basis of either configural or featural modifications was measured under the same four conditions. Although the ability to discriminate configural modifications was superior in the medium-frequency condition, so was the ability to discriminate featural modifications. We conclude that the band of SF that is critical for face recognition does not contribute preferentially to configural encoding.     

A feature-inversion effect: can an isolated feature show behavior like the face-inversion effect?

The face-inversion effect (FIE) is explained by the configural-processing hypothesis. It proposes that inversion disrupts configural information processing (spatial links among facial features) and leaves the processing of featural information (eyes, nose, and mouth) comparatively intact. According to this hypothesis, an inverted isolated facial feature cannot show a feature-inversion effect--that is, behavior similar to the FIE--since all the spatial links between it and the other features in a face are eliminated; that is, the configural information is removed. The findings of the present study, which show that isolated eyes do exhibit the feature-inversion effect, support the extended configural-processing hypothesis. This proposes that inversion also impairs processing of the configural information within the eyes themselves. Removal of the brows in whole faces tended to interfere with processing of the configural information in the upright position but to facilitate processing in the inverted position.     

Picture-plane inversion leads to qualitative changes of face perception

Presenting a face stimulus upside-down generally causes a larger deficit in perceiving metric distances between facial features ("configuration") than local properties of these features. This effect supports a qualitative account of face inversion: the same transformation affects the processing of different kinds of information differently. However, this view has been recently challenged by studies reporting equal inversion costs of performance for discriminating featural and configural manipulations on faces. In this paper I argue that these studies did not replicate previous results due to methodological factors rather than largely irrelevant parameters such as having equal performance for configural and featural conditions at upright orientation, or randomizing trials across conditions. I also argue that identifying similar diagnostic features (eyes and eyebrows) for discriminating individual faces at upright and inverted orientations by means of response classification methods does not dismiss at all the qualitative view of face inversion. Considering these elements as well as both behavioral and neuropsychological evidence, I propose that the generally larger effect of inversion for processing configural than featural cues is a mere consequence of the disruption of holistic face perception. That is, configural relations necessarily involve two or more distant features on the face, such that their perception is most dependent on the ability to perceive simultaneously multiple features of a face as a whole.     

Features are also important: contributions of featural and configural processing to face recognition

It has been suggested that face recognition is primarily based on configural information, with featural information playing little or no role. We investigated this idea by comparing the prototype effect for face prototypes that emphasized either featural or configural processing. In Experiment 1, participants showed a tendency to commit false alarms in response to nonstudied prototypes, and this tendency was equivalent for featural and configural prototypes. Experiment 2 replicated this finding, and provided support for the assumption that the two types of prototypes differed in terms of featural and configural processing: Face inversion eliminated the prototype effect for configural prototypes but not for featural prototypes. These results suggest that both featural and configural processing make important contributions to face recognition, and that their effects are dissociable.     

When feature information comes first! Early processing of inverted faces

We investigated the early stages of face recognition and the role of featural and holistic face information. We exploited the fact that, on inversion, the alienating disorientation of the eyes and mouth in thatcherised faces is hardly detectable. This effect allows featural and holistic information to be dissociated and was used to test specific face-processing hypotheses. In inverted thatcherised faces, the cardinal features are already correctly oriented, whereas in undistorted faces, the whole Gestalt is coherent but all information is disoriented. Experiment 1 and experiment 3 revealed that, for inverted faces, featural information processing precedes holistic information. Moreover, the processing of contextual information is necessary to process local featural information within a short presentation time (26 ms). Furthermore, for upright faces, holistic information seems to be available faster than for inverted faces (experiment 2). These differences in processing inverted and upright faces presumably cause the differential importance of featural and holistic information for inverted and upright faces.     

Emotion recognition: The role of featural and configural face information

Several studies investigated the role of featural and configural information when processing facial identity. A lot less is known about their contribution to emotion recognition. In this study, we addressed this issue by inducing either a featural or a configural processing strategy (Experiment 1) and by investigating the attentional strategies in response to emotional expressions (Experiment 2). In Experiment 1, participants identified emotional expressions in faces that were presented in three different versions (intact, blurred, and scrambled) and in two orientations (upright and inverted). Blurred faces contain mainly configural information, and scrambled faces contain mainly featural information. Inversion is known to selectively hinder configural processing. Analyses of the discriminability measure (A′) and response times (RTs) revealed that configural processing plays a more prominent role in expression recognition than featural processing, but their relative contribution varies depending on the emotion. In Experiment 2, we qualified these differences between emotions by investigating the relative importance of specific features by means of eye movements. Participants had to match intact expressions with the emotional cues that preceded the stimulus. The analysis of eye movements confirmed that the recognition of different emotions rely on different types of information. While the mouth is important for the detection of happiness and fear, the eyes are more relevant for anger, fear, and sadness.     

Configural and featural processing during face perception: A new stimulus set

A new stimulus set of 60 male-face stimuli in seven in-depth orientations was developed. The set can be used in research on configural versus featural mechanisms of face processing. Configural, or holistic, changes are produced by changing the global form of the face, whereas featural, or part-based, changes are attained by altering the local form of internal facial features. For each face in the set, there is one other face that differs only by its global form and one other face that differs only by its internal features. In all faces, extrafacial cues have been eliminated or standardized. The stimulus set also contains a color-coded division of each face in areas of interest, which is useful for eye movement research on face scanning strategies. We report a matching experiment with upright and inverted face pairs that demonstrates that the face stimulus set is indeed useful for research on configural and featural face perception. The stimulus set may be downloaded from the Psychonomic Society’s archive (brm.psychonomic-journals.org/content/supplemental) or from our Web site (http://ppw.kuleuven .be/labexppsy/ newSite/resources)     

A sideways look at configural encoding: two different effects of face rotation

Inversion has a disproportionate disruptive effect on the recognition of faces. This may be due to the disruption of holistic or configural encoding employed to recognise upright faces. The paradigm developed by Tanaka and Farah (1993 Quarterly Journal of Experimental Psychology, Section A 46 225-246) was used to investigate the effect of 90 degrees (or orthogonal) rotation on configural encoding. Faces learnt in the orthogonal condition were not recognised as well as upright faces, but a whole-face advantage was found in both cases. This whole-face advantage did not occur for inverted faces. It appears that 90 degrees rotation affects recognition but not specifically configural encoding. It is concluded that rotating a face can have at least two different effects on face processing depending on the range of rotation. Implications for the nature of facial dimensions and the expertise account of the inversion effect are considered.     

Perception of novel faces: the parts have it!

It has been suggested that, as a result of expertise, configural information plays a predominant role in face processing. We investigated this idea using novel and learned faces. In experiment 1, sixteen participants matched two subsequently presented blurred or scrambled faces, which could be either upright or inverted, in a sequential same -different matching task. By means of blurring, featural information is hampered, whilst scrambling disrupts configural information. Each face was unfamiliar to the participants and was presented for 1000 ms. An ANOVA on the d' values revealed a significant advantage for scrambled faces. In experiment 2, fourteen participants were tested with the same design, except that the second face was always intact. Again, the ANOVA on the d' values revealed a significant advantage for scrambled faces. In experiment 3 half of the faces were learned in a familiarisation block prior to the experiment. The ANOVA of these d' values revealed a significant interaction of familiarity and condition, showing that blurred stimuli were better recognised when the faces were familiar. These results suggest that recognition of novel faces, compared to learned faces, relies relatively more on the processing of featural information. In the course of familiarisation the importance of configural information increases.     

面孔加工的结构编码阶段异族效应的加工方式： ERPs研究

实验采用事件相关电位（ERPs）技术,以大学生为被试,运用非洲、欧洲、亚洲三个不同种族的面孔照片,探究异族效应与加工方式的关系。行为实验结果显示,被试对非洲面孔的反应时快于欧洲面孔,且两者均快于亚洲面孔。ERP研究结果显示,亚洲面孔的N170波幅显著小于非洲和欧洲面孔的波幅,后两者的N170波幅没有差异,而且亚洲面孔的倒置效应最大。研究结果说明相对于异族面孔,本族面孔更多地使用构型编码的加工方式,且N170成分表现出右半球优势。     

Multiple perceptual strategies used by macaque monkeys for face recognition

Successful integration of individuals in macaque societies suggests that monkeys use fast and efficient perceptual mechanisms to discriminate between conspecifics. Humans and great apes use primarily holistic and configural, but also feature-based, processing for face recognition. The relative contribution of these processes to face recognition in monkeys is not known. We measured face recognition in three monkeys performing a visual paired comparison task. Monkey and humans faces were (1) axially rotated, (2) inverted, (3) high-pass filtered, and (4) low-pass filtered to isolate different face processing strategies. The amount of time spent looking at the eyes, mouth, and other facial features was compared across monkey and human faces for each type of stimulus manipulation. For all monkeys, face recognition, expressed as novelty preference, was intact for monkey faces that were axially rotated or spatially filtered and was supported in general by preferential looking at the eyes, but was impaired for inverted faces in two of the three monkeys. Axially rotated, upright human faces with a full range of spatial frequencies were also recognized, however, the distribution of time spent exploring each facial feature was significantly different compared to monkey faces. No novelty preference, and hence no inferred recognition, was observed for inverted or low-pass filtered human faces. High-pass filtered human faces were recognized, however, the looking pattern on facial features deviated from the pattern observed for monkey faces. Taken together these results indicate large differences in recognition success and in perceptual strategies used by monkeys to recognize humans versus conspecifics. Monkeys use both second-order configural and feature-based processing to recognize the faces of conspecifics, but they use primarily feature-based strategies to recognize human faces.