Not just the norm: Exemplar-based models also predict face aftereffects

The face recognition literature has considered two competing accounts of how faces are represented within the visual system: Exemplar-based models assume that faces are represented via their similarity to exemplars of previously experienced faces, while norm-based models assume that faces are represented with respect to their deviation from an average face, or norm. Face identity aftereffects have been taken as compelling evidence in favor of a norm-based account over an exemplar-based account. After a relatively brief period of adaptation to an adaptor face, the perceived identity of a test face is shifted toward a face with attributes opposite to those of the adaptor, suggesting an explicit psychological representation of the norm. Surprisingly, despite near universal recognition that face identity aftereffects imply norm-based coding, there have been no published attempts to simulate the predictions of norm- and exemplar-based models in face adaptation paradigms. Here, we implemented and tested variations of norm and exemplar models. Contrary to common claims, our simulations revealed that both an exemplar-based model and a version of a two-pool norm-based model, but not a traditional norm-based model, predict face identity aftereffects following face adaptation.     

Identification and ratings of caricatures: Implications for mental representations of faces

S. Brennan (1985, Leonardo, 18, 170–178) has developed a computer-implemented caricature generator based on a holistic theory of caricature. A face is represented by 37 lines, based on a fixed set of 169 points. Caricatures are produced by exaggerating all metric differences between a face and a norm. Anticaricatures can be created by reducing all the differences between a face and a norm. Caricatures of familiar faces were identified more quickly than veridical line drawings, which were identified more quickly than anticaricatures. There was no difference in identification accuracy for the three types of representation. The best likeness was considered to be a caricature. We discuss the implications of these results for how faces are mentally represented. The results are consistent with a holistic theory of encoding in which distinctive aspects of a face are represented by comparison with a norm. We suggest that this theory may be appropriate for classes of visual stimuli, other than faces, whose members share a configuration definable by a fixed set of points.     

The effect of educational environment on identity recognition and perceptions of within-person variability

Individuals from small communities show impoverished face recognition relative to those from large communities, suggesting that the number of faces to which one is exposed has a measurable effect on face processing abilities. We sought to extend these findings by examining a second factor that influences the population of faces to which one is exposed during childhood: educational setting. In particular, we examined whether formerly home-schooled participants show reduced performance relative to non-homeschoolers on the Cambridge Face Memory Test (CFMT) and on a sorting task in which participants sort photographs of two unfamiliar identities into piles representing the number of identities they believe are present. On the CFMT, there was no effect of educational setting. However, formerly home-schooled participants showed significant deficits on the sorting task. Such results suggest that reduced exposure to faces early in life as a function of home-schooling may have lasting effects on the face processing system.     

Other-race face perception

The other-race effect (or own-race bias or cross-racial identification effect) refers to the finding that recognition memory tends to be better for faces of members of subjects' own race than for faces of members of other races. The current study was designed to test the hypothesis that perceptual skills specific to identifying faces of particular racial groups contribute to this effect. On each of 50 trials, a photograph of a face was tachistoscopically presented for 120 ms, followed by a pattern mask and then a plain-view test pair composed of the previously presented face and a matched foil. As predicted, an other-race effect was obtained on this perceptual task: White subjects performed significantly more poorly on trials involving African American faces than on trials involving White faces, whereas no such difference was obtained among African American subjects.     

Effects of perceptual quality on the processing of human faces presented to the left and right cerebral hemispheres

Three experiments examined the effects of stimulus duration, retinal eccentricity, and visual noise on the processing of human faces presented to the left visual field/right hemisphere (LVF-RH) and right visual field/left hemisphere (RVF-LH). In Experiment 1 observers identified which of 10 similar male faces was presented on a screen. The single face was presented for 10, 55, or 100 ms at 1 degree, 4 degrees, or 9 degrees of visual angle to the left or right of fixation. Decreasing stimulus duration and increasing retinal eccentricity lowered face recognition. The effect of duration was the same for LVF-RH and RVF-LH trials, but the detrimental effect of increasing retinal eccentricity was larger on LVF-RH trials than on RVF-LH trials. In Experiment 2 observers indicated whether a single face from this same set was a member of a memorized set of five positive faces. The probe face on each trial was presented alone or embedded in visual noise. Visual noise increased the error rate more on LVF-RH trials than on RVF-LH trials. This effect was replicated in Experiment 3, which also required observers to make a much easier discrimination between male and female faces. In the male/female task visual noise tended to impair performance more on RVF-LH trials than on LVF-RH trials, opposite the effect for the male/male task. These results are discussed in terms of hemispheric asymmetry for global versus local features of faces, the level of feature analysis demanded by a task, and the level of feature analysis most disrupted by perceptual degradation.     

Early detection and avoidance of threatening faces during passive viewing

To evaluate whether there is an early attentional bias towards negative stimuli, we tracked participants' eyes while they passively viewed displays composed of four Ekman faces. In Experiment 1 each display consisted of three neutral faces and one face depicting fear or happiness. In half of the trials, all faces were inverted. Although the passive viewing task should have been very sensitive to attentional biases, we found no evidence that overt attention was biased towards fearful faces. Instead, people tended to actively avoid looking at the fearful face. This avoidance was evident very early in scene viewing, suggesting that the threat associated with the faces was evaluated rapidly. Experiment 2 replicated this effect and extended it to angry faces. In sum, our data suggest that negative facial expressions are rapidly analysed and influence visual scanning, but, rather than attract attention, such faces are actively avoided.     

On the other side of the mean: the perception of dissimilarity in human faces

We created a 'face space' using a laser-scan representation of faces. In this space, a caricature can be made by moving a face away from the average face, along the line connecting the particular face to the average face. Here, we move the face along this line in the other direction, proceeding through the mean and 'out the other side'. This results in a face that is 'opposite', in a computational sense, to the original face. We morphed several faces into their anti-faces and sampled the morph trajectory in five discrete steps. We then collected similarity ratings from human participants for all possible pairs of morphed faces to determine how the distances in the 'physical face space' related to the distances in the 'psychological face space'. The data indicate that there is a perceptual discontinuity of face identity as the face crosses over to the 'other side of the mean'. We consider these results in the context of face-space models of human face processing.     

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。     

The effect of inversion on the encoding of normal and “thatcherized” faces

In the “Thatcher illusion” a face, in which the eyes and mouth are inverted relative to the rest of the face, looks grotesque when shown upright but not when inverted. In four experiments we investigated the contribution of local and global processing to this illusion in normal observers. We examined inversion effects (i.e., better performance for upright than for inverted faces) in a task requiring discrimination of whether faces were or were not “thatcherized”. Observers made same/different judgements to isolated face parts (Experiments 1-2) and to whole faces (Experiments 3-4). Face pairs had the same or different identity, allowing for different process- ing strategies using feature-based or configural information, respectively. In Experiment 1, feature-based matching of same-person face parts yielded only a small inversion effect for normal face parts. However, when feature-based matching was prevented by using the face parts of different people on all trials (Experiment 2) an inversion effect occurred for normal but not for thatcherized parts. In Experiments 3 and 4, inversion effects occurred with normal but not with thatcherized whole faces, on both same- and different-person matching tasks. This suggests that a common configural strategy was used with whole (normal) faces. Face context facilitated attention to misoriented parts in same-person but not in different-person matching. The results indicate that (1) face inversion disrupts local configural processing, but not the processing of image features, and (2) thatcherization disrupts local configural processing in upright faces.     

AVERAGENESS, EXAGGERATION, AND FACIAL ATTRACTIVENESS

Abstract— Langlois and her colleagues reported in this journal that composite faces are more attractive than the component faces used to create them, and conjectured that averageness is attractive (Langlois & Roggman, 1990, Langlois, Roggman, & Musselman, 1994) However, extremes may also be attractive (Perrett, May, & Yoshikawa, 1994). We investigated the effect of averageness (proximity to a norm or average face) on attractiveness using a computerized caricature generator to very averageness. Attractiveness increased with averageness (Experiment 1) and was negatively correlated with distinctiveness, a subjective measure of the converse of averageness (Experiments 1 and 2) Extremes (caricatures) were not attractive Line-drawing composites, which avoid some of the problems associated with gray-level composites, were significantly more attractive and less distinctive (more average) than individual faces (Experiment 2) These results support the claim that averageness is attractive.     

同龄效应的类别化-个体化模型

针对面孔认知偏好的类别化-个体化模型,探讨该模型中的知觉经验、加工方式和动机水平对本龄效应的影响。结果表明增加对异龄面孔的知觉经验、个体化加工异龄面孔及提高加工异龄面孔的动机都能有效改善对异龄面孔的再认,减少本龄效应量。这说明本龄效应遵循类别化-个体化模型,且在类别化-个体化模型中本龄效应与本族效应加工原理相仿。。     

Electrophysiological correlates of the composite face illusion: Disentangling perceptual and decisional components of holistic face processing in the human brain

When the bottom halves of two faces differ, people’s behavioral judgment of the identical top halves of those faces is impaired: they report that the top halves are different, and/or take more time than usual to provide a response. This behavioral measure is known as the composite face effect (CFE) and has traditionally been taken as evidence that faces are perceived holistically. Recently, however, it has been claimed that this effect is driven almost entirely by decisional, rather than perceptual, factors ( Richler, Gauthier, Wenger, & Palmeri, 2008). To disentangle the contribution of perceptual and decisional brain processes, we aimed to obtain an event-related potential (ERP) measure of the CFE at a stage of face encoding ( Jacques & Rossion, 2009) in the absence of a behavioral CFE effect. Sixteen participants performed a go/no-go task in an oddball paradigm, lifting a finger of their right or left hand when the top half of a face changed identity. This change of identity of the top of the face was associated with an increased ERP signal on occipito-temporal electrode sites at the N170 face-sensitive component (∼160 ms), the later decisional P3b component, and the lateralized readiness potential (LRP) starting at ∼350 ms. The N170 effect was observed equally early when only the unattended bottom part of the face changed, indicating that an identity change was perceived across the whole face in this condition. Importantly, there was no behavioral response bias for the bottom change trials, and no evidence of decisional biases from electrophysiological data (no P3b and LRP deflection in no-go trials). These data show that an early CFE can be measured in ERPs in the absence of any decisional response bias, indicating that the CFE reflects primarily the visual perception of the whole face.     

Reduced adaptability,but no fundamental disruption,of norm‐based face coding following early visual deprivation from congenital cataracts

Faces are adaptively coded relative to visual norms that are updated by experience, and this adaptive coding is linked to face recognition ability. Here we investigated whether adaptive coding of faces is disrupted in individuals (adolescents and adults) who experience face recognition difficulties following visual deprivation from congenital cataracts in infancy. We measured adaptive coding using face identity aftereffects, where smaller aftereffects indicate less adaptive updating of face‐coding mechanisms by experience. We also examined whether the aftereffects increase with adaptor identity strength, consistent with norm‐based coding of identity, as in typical populations, or whether they show a different pattern indicating some more fundamental disruption of face‐coding mechanisms. Cataract‐reversal patients showed significantly smaller face identity aftereffects than did controls (Experiments 1 and 2). However, their aftereffects increased significantly with adaptor strength, consistent with norm‐based coding (Experiment 2). Thus we found reduced adaptability but no fundamental disruption of norm‐based face‐coding mechanisms in cataract‐reversal patients. Our results suggest that early visual experience is important for the normal development of adaptive face‐coding mechanisms.     

The dissociable influence of social context on judgements of facial attractiveness and trustworthiness

The cheerleader effect occurs when the same face is rated to be more attractive when it is seen in a group compared to when seen alone. We investigated whether this phenomenon also occurs for trustworthiness judgements, and examined how these effects are influenced by the characteristics of the individual being evaluated and those of the group they are seen in. Across three experiments, we reliably replicated the cheerleader effect. Most faces became more attractive in a group. Yet, the size of the cheerleader effect that each face experienced was not related to its own attractiveness, nor to the attractiveness of the group or the group’s digitally averaged face. We discuss the implications of our findings for the hierarchical encoding and contrast mechanisms that have previously been used to explain the cheerleader effect. Surprisingly, judgements of facial trustworthiness did not experience a ‘cheerleader effect’. Instead, we found that untrustworthy faces became significantly more trustworthy in all groups, while there was no change for faces that were already trustworthy alone. Taken together, our results demonstrate that social context can have a dissociable influence on our first impressions, depending on the trait being evaluated.     

Internal reinstatement hides cuing effects in source memory tasks

Reinstating source details at test often has no impact on source memory. We tested the proposition that participants internally reinstate source cues when such cues are not provided by the experimenter, thus making the external cues redundant. Participants studied words paired with either a male or a female face and were later asked to specify the gender of the face studied with each word. To disrupt the ability to internally reinstate sources, some participants saw eight male faces and eight female faces throughout the study list (multiple-face condition), making it difficult to determine which face should be internally reinstated for uncued test trials. Other participants saw only a single face for each gender (single-face condition), which should facilitate internal reinstatement. Across three experiments, participants in the multiple-face condition showed improved source discrimination when the studied faces were reinstated at test, as compared to uncued trials. In contrast, participants in the single-face condition showed no effect of the face cues. Moreover, the cuing effect for the multiple-face condition disappeared when the test structure facilitated internal reinstatement. Overall, the experiments support the contention that internal reinstatement is a natural part of source retrieval that can mask the effects of external cues.     

Fitting the child's mind to the world: adaptive norm-based coding of facial identity in 8-year-olds

In adults, facial identity is coded by opponent processes relative to an average face or norm, as evidenced by the face identity aftereffect: adapting to a face biases perception towards the opposite identity, so that a previously neutral face (e.g. the average) resembles the identity of the computationally opposite face. We investigated whether children as young as 8 use adaptive norm-based coding to represent faces, a question of interest because 8-year-olds are less accurate than adults at recognizing faces and do not show the adult neural markers of face expertise. We found comparable face identity aftereffects in 8-year-olds and adults: perception of identity in both groups shifted in the direction predicted by norm-based coding. This finding suggests that, by 8 years of age, the adaptive computational mechanisms used to code facial identity are like those of adults and hence that children's immaturities in face processing arise from another source.     

The time course of processing external and internal features of unfamiliar faces

The time course of processing internal and external facial features was studied in a sequential face matching task, where first a target face was presented, followed by a test face. The exposure duration of the test face was varied systematically (90, 120, 150 ms, and self-paced). In three tasks, participants were instructed to match either the whole face, only external features, or only internal features of the target and test face. Taken together, the results in all the three tasks provide evidence for very fast matching processes. For upright faces, maximal performance was achieved at 90 ms exposure duration and longer exposure durations (120, 150 ms, self-paced) did not improve accuracy. For inverted whole faces, reduced exposure duration resulted in an increase of matching errors, suggesting that below 150 ms of exposure duration, inverted faces cannot be matched reliably. When matching selected facial features only, no such inversion effect was found. Our data challenges previous claims that external features are matched faster than internal: no difference of time course was found between external and internal features. However, external features were matched more accurately.     

Bidirectional feature selection with global and local structure preservation for small size samples

Collecting samples is a challenging task for face recognition, especially for some real-world applications such as law enhancement and ID card identification, where there is usually single sample per person (SSPS) used to train a face recognition system. To extract discriminative features from the small size samples, in this paper we propose virtual samples via bidirectional feature selection with global and local structure preservation (VS-BFS-GL) to augment the number of training samples. In VS-BFS-GL, bidirectional feature selection is developed, which introduces L_2,1 norm to explore the face variations from both horizontal and vertical directions. Further, to include more variations in the virtual images, the global structure information and sample-specified local structure information of the SSPP training set are considered. By integrating bidirectional feature selection, global and local structure, the limited training samples are fully utilized and more knowledge are mined. To further improve the effectiveness of VS-BFS-GL, an auxiliary database containing different face variations can be used to explore the local structure information. We extensively evaluated the proposed approach on AR and FERET database. The promising recognition results demonstrate that VS-BFS-GL is robust to expression, pose and partial occlusion variations in the faces.     

Symmetry, averageness, and feature size in the facial attractiveness of women

Female facial attractiveness was investigated by comparing the ratings made by male judges with the metric characteristics of female faces. Three kinds of facial characteristics were considered: facial symmetry, averageness, and size of individual features. The results suggested that female face attractiveness is greater when the face is symmetrical, is close to the average, and has certain features (e.g., large eyes, prominent cheekbones, thick lips, thin eyebrows, and a small nose and chin). Nevertheless, the detrimental effect of asymmetry appears to result solely from the fact that an asymmetrical face is a face that deviates from the norm. In addition, a factor analysis indicated that averageness best accounts for female attractiveness, but certain specific features can also be enhancing.