Holistic processing of shape cues in face identification: Evidence from face inversion,composite faces,and acquired prosopagnosia

According to a classical functional architecture of face processing (Bruce & Young, 1986), sex processing on faces is a parallel function to individual face recognition. One consequence of the model is thus that sex categorization on faces is not influenced by face familiarity. However, the behavioural and neuro-psychological evidences supporting this dissociation are yet equivocal. To test the independence between sex processing on faces and familiar face recognition, familiar (learned) faces were morphed with new faces, generating facial continua of visual similarity to familiar faces. First, a pilot experiment shown that subjects familiarized with one extreme of the face continuum roughly perceive one half of the continuum (60 to 100% of visual similarity to familiar faces) as made of familiar faces and the other part as unfamiliar. In the experiment proper, subjects were familiarized with faces and tested in a sex decision task made on faces at the different steps of the continua. Subjects were significantly quicker at telling the sex of faces perceived as familiar (60-100%), and the effect was not observed in a control (untrained) group. These results indicate that familiar face representations are activated before sex categorization is completed, and can facilitate this processing. The nature of the interaction between sex categorization on faces and familiar face recognition is discussed.     

Recognising the ageing face: the role of age in face processing.

The effect of age-induced changes on face recognition were investigated as a means of exploring the role of age in the encoding of new facial memories. The ability of participants to recognise each of six previously learnt faces was tested with versions which were either identical to the learnt faces, the same age (but different in pose and expression), or younger or older in age. Participants were able to cope well with facial changes induced by ageing: their performance with older, but not younger, versions was comparable to that with faces which differed only in pose and expression. Since the large majority of different age versions were recognised successfully, it can be concluded that the process of recognition does not require an exact match in age characteristics between the stored representation of a face and the face currently in view. As the age-related changes explored here were those that occur during the period of growth, this in turn implies that the underlying structural physical properties of the face are (in addition to pose and facial expression) invariant to a certain extent.     

Objects that induce face pareidolia are prioritized by the visual system

The human visual system has evolved specialized neural mechanisms to rapidly detect faces. Its broad tuning for facial features is thought to underlie the illusory perception of faces in inanimate objects, a phenomenon called face pareidolia. Recent studies on face pareidolia suggest that the mechanisms underlying face processing, at least at the early stages of visual encoding, may treat objects that resemble faces as real faces; prioritizing their detection. In our study, we used breaking continuous flash suppression (b-CFS) to examine whether the human visual system prioritizes the detection of objects that induce face pareidolia over stimuli matched for object content. Similar to previous b-CFS results using real face stimuli, we found that participants detected the objects with pareidolia faces faster than object-matched control stimuli. Given that face pareidolia has been more frequently reported amongst individuals prone to hallucinations, we also explored whether this rapid prioritization is intact in individuals with schizophrenia, and found evidence suggesting that it was. Our findings suggest that face pareidolia engages a broadly tuned mechanism that facilitates rapid face detection. This may involve the proposed fast subcortical pathway that operates outside of visual awareness.     

Can holistic processing be learned for inverted faces?

The origin of "special" processing for upright faces has been a matter of ongoing debate. If it is due to generic expertise, as opposed to having some innate component, holistic processing should be learnable for stimuli other than upright faces. Here we assess inverted faces. We trained subjects to discriminate identical twins using up to 1100 exposures to each twin in different poses and images. In the upright orientation, twin discrimination was supported by holistic processing. Removal of a single face feature had no effect on performance, and a composite effect (Young, A. W., Hellawell, D., & Hay, D.C. (1987). Configurational information in face perception. Perception 16 (6), 747-759) was obtained. In the inverted orientation, however, above chance identification ability relied on (a) image specific learning, or (b) tiny local feature differences not noticed in the upright faces. The failure to learn holistic processing for inverted faces indicates that, in contrast to the situation for objects (Tarr, M.J., & Pinker, S. (1989). Mental rotation and orientation-dependence in shape recognition. Cognitive Psychology 21 (2), 233-282), orientation specificity of face processing is highly stable against practice.     

熟悉面孔与陌生面孔识别的差异

摘 要：面孔识别包括熟悉面孔识别和陌生面孔识别,两者之间的差异已经得到了多方面的证明。神经生理学的研究发现,识别陌生面孔和熟悉面孔所激活的脑区存在差异,与识别陌生面孔相比,识别熟悉面孔的脑区活动范围更大,包括双侧前额叶、单侧颞叶、海马回、杏仁核、后扣带回和右下额叶。并且也存在脑电的差异,比如 N400和P600的差异。对陌生面孔和熟悉面孔识别具有相同影响的因素有：遮挡、倒置等;具有不同影响的因素有：视角、表情、内外部特征等。文章也探讨了陌生面孔变成熟悉面孔的过程,以及在这一过程中起主要作用的编码形式等。     

Effect of partial occlusion on newborns' face preference and recognition

Many studies have shown that newborns prefer (e.g. Goren, Sarty & Wu, 1975; Valenza, Simion, Macchi Cassia & Umiltà, 1996) and recognize (e.g. Bushnell, Say & Mullin, 1989; Pascalis & de Schonen, 1994) faces. However, it is not known whether, at birth, faces are still preferred and recognized when some of their parts are not visible because hindered by other configurations, that is when faces are partly occluded. Also, it is not known whether newborns' preference for an upright over an inverted face and newborns' face recognition are differentially affected depending on the salience of the occluded face features. Seventy-seven newborns (mean age of 43.5 hrs) were tested using the preferential looking (Experiment 1) and the habituation techniques (Experiment 2). Results demonstrated that newborns prefer and recognize occluded faces even if some portions of them are not available, at least when the hindered features are not salient. On the contrary, these abilities are affected by obscuring high salience facial features (i.e. eyes). However, while in the case of face detection, eyes occlusion completely prevented newborns' face detection, in the case of face recognition an analogous stimulus manipulation heavily impaired, but did not totally preclude, newborns' recognition performance. The data collected improve our comprehension of newborns' way of processing and encoding information to detect and recognize faces.     

Finding an unfamiliar face in a line-up: Viewing multiple images of the target is beneficial on target-present trials but costly on target-absent trials

When viewing unfamiliar faces, photographs of the same person often are perceived as belonging to different people and photographs of different people as belonging to the same person. Identity matching of unfamiliar faces is especially challenging when the photographs are of a person whose ethnicity differs from that of the observer. In contrast, matching is trivial when viewing familiar faces, regardless of race. Viewing multiple images of an own-race target identity improves accuracy on a line-up task when the target is known to be present (Dowsett et al., 2016, Q J Exp Psychol, 69, 1), suggesting that exposure to within-person variability in appearance is key to face learning. Across three experiments, we show that viewing multiple images of a target identity also improves accuracy for other-race faces on target-present trials. However, viewing multiple images decreases accuracy (i.e., increases false alarms) on target-absent trials for both own- and other-race faces. We discuss the implications of our findings for models of face recognition and for forensic settings.     

Face detection dissociates from face identification

We describe three experiments in which viewers complete face detection tasks as well as standard measures of unfamiliar face identification. In the first two studies, participants viewed pareidolic images of objects (Experiment 1) or cloud scenes (Experiment 2), and their propensity to see faces in these scenes was measured. In neither case is performance significantly associated with identification, as measured by the Cambridge Face Memory or Glasgow Face Matching Tests. In Experiment 3 we showed participants real faces in cluttered scenes. Viewers’ ability to detect these faces is unrelated to their identification performance. We conclude that face detection dissociates from face identification.     

Perceptual learning in face processing: Comparison facilitates face recognition

Theoretical analyses of perceptual learning suggest that comparison between similar stimuli aids subsequent discrimination between them. The current studies examined whether the opportunity to compare a target face to other similar faces during a preexposure phase facilitated performance on a matching task. Performance was better when the target face had been presented in alternation with similar comparator faces than if that target had not been exposed before test. Exposure to the target face without comparators, or exposure alternating with dissimilar faces, improved performance relative to the nonexposed control but performance was not as good as when target exposure was given in alternation to similar comparison faces. The effects were not influenced by image changes between exposure and test. It is suggested that exposure to a face in comparison to similar stimuli focuses the central representation of a face on its unique features. In practical terms these results suggest that reliable identification of an individual from their photograph would be improved by viewing that photograph in comparison with photographs of similar people.     

Ingroup and outgroup differences in face detection

Humans show improved recognition for faces from their own social group relative to faces from another social group. Yet before faces can be recognized, they must first be detected in the visual field. Here, we tested whether humans also show an ingroup bias at the earliest stage of face processing – the point at which the presence of a face is first detected. To this end, we measured viewers' ability to detect ingroup (Black and White) and outgroup faces (Asian, Black, and White) in everyday scenes. Ingroup faces were detected with greater speed and accuracy relative to outgroup faces (Experiment 1). Removing face hue impaired detection generally, but the ingroup detection advantage was undiminished (Experiment 2). This same pattern was replicated by a detection algorithm using face templates derived from human data (Experiment 3). These findings demonstrate that the established ingroup bias in face processing can extend to the early process of detection. This effect is ‘colour blind’, in the sense that group membership effects are independent of general effects of image hue. Moreover, it can be captured by tuning visual templates to reflect the statistics of observers' social experience. We conclude that group bias in face detection is both a visual and a social phenomenon.     

Repetition priming of face gender judgments: An instance based explanation

Earlier studies of repetition priming using faces have been interpreted as indicating that such effects are confined to the processing of known faces. The experiment reported here employed eight rather than the more usual two presentation trials and required subjects to make gender decisions (is it a male or is it a female face?) to both familiar and unfamiliar faces. This allowed the currently favored recognition unit theories of face processing to be compared with the Logan (1988) instance model. Equivalent repetition priming effects were observed for both familiar and unfamiliar faces and were well fitted by power functions. It is argued that the findings are consistent with the strong predictions made by Logan's model and pose problems for recognition unit based theories.     

The effects of social anxiety on emotional face discrimination and its modulation by mouth salience

People high in social anxiety experience fear of social situations due to the likelihood of social evaluation. Whereas happy faces are generally processed very quickly, this effect is impaired by high social anxiety. Mouth regions are implicated during emotional face processing, therefore differences in mouth salience might affect how social anxiety relates to emotional face discrimination. We designed an emotional facial expression recognition task to reveal how varying levels of sub-clinical social anxiety (measured by questionnaire) related to the discrimination of happy and fearful faces, and of happy and angry faces. We also categorised the facial expressions by the salience of the mouth region (i.e. high [open mouth] vs. low [closed mouth]). In a sample of 90 participants higher social anxiety (relative to lower social anxiety) was associated with a reduced happy face reaction time advantage. However, this effect was mainly driven by the faces with less salient closed mouths. Our results are consistent with theories of anxiety that incorporate an oversensitive valence evaluation system.     

An ERP study of famous face incongruity detection in middle age

Age-related changes in famous face incongruity detection were examined in middle-aged (mean = 50.6) and young (mean = 24.8) subjects. Behavioral and ERP responses were recorded while subjects, after a presentation of a "prime face" (a famous person with the eyes masked), had to decide whether the following "test face" was completed with its authentic eyes (congruent) or with other eyes (incongruent). The principal effects of advancing age were (1) behavioral difficulties in discriminating between incongruent and congruent faces; (2) a reduced N400 effect due to N400 enhancement for both congruent and incongruent faces; (3) a latency increase of both N400 and P600 components. ERPs to primes (face encoding) were not affected by aging. These results are interpreted in terms of early signs of aging.     

Repetition priming of face gender judgments: An instance based explanation

Earlier studies of repetition priming using faces have been interpreted as indicating that such effects are confined to the processing of known faces. The experiment reported here employed eight rather than the more usual two presentation trials and required subjects to make gender decisions (is it a male or is it a female face?) to both familiar and unfamiliar faces. This allowed the currently favored recognition unit theories of face processing to be compared with the Logan (1988) instance model. Equivalent repetition priming effects were observed for both familiar and unfamiliar faces and were well fitted by power functions. It is argued that the findings are consistent with the strong predictions made by Logan's model and pose problems for recognition unit based theories.     

Recognizing moving faces: a psychological and neural synthesis

Information for identifying a human face can be found both in the invariant structure of features and in idiosyncratic movements and gestures. When both kinds of information are available, psychological evidence indicates that: (1) dynamic information contributes more to recognition under non-optimal viewing conditions, e.g. poor illumination, low image resolution, recognition from a distance; (2) dynamic information contributes more as a viewer's experience with the face increases; and (3) a structure-from-motion analysis can make a perceptually based contribution to face recognition. A recently proposed distributed neural system for face perception, with minor modifications, can accommodate the psychological findings with moving faces.     

Still another confounded face in the crowd

Experiments using schematic faces developed by Öhman (Öhman, Lundqvist, &; Esteves, 2001) seem to document an anger-superiority effect, although we have come to question these experiments. Our work shows that the low-level features of these schematic faces interact with the face’s surround to produce effects that have been attributed to facial affect. Using relatively neutral faces that preserved the feature and surround spatial relationships of angry and happy schematic faces, we produced reaction times (RTs) that were indistinguishable from those found with angry and happy faces. We also found that the target face’s position within the crowd determined the magnitude of the advantage for angry faces as well as for relatively affect-neutral faces. Removing the facial surround reduces the advantage for angry faces, largely by improving performance on happy faces. There was an apparent small advantage for angry features without a surround. öhman faces avoid the problems associated with modified grayscale faces only to introduce an equally troubling confound.     

The role of color in human face detection

Significant advances have been made in understanding human face recognition . However, a fundamental aspect of this process, how faces are located in our visual environment, is poorly understood and little studied. Here we examine the role of color in human face detection. We demonstrate that detection performance declines when color information is removed from faces, regardless of whether the surrounding scene context is rendered in color. Furthermore, faces rendered in unnatural colors are hard to detect, suggesting a role beyond simple segmentation. When faces are presented such that half the surface is colored appropriately, and half unnaturally, performance declines. This suggests that observers are not simply using the presence of skin color "patches" to detect faces. Rather, our data suggest that detection operates via a face template combining diagnostic color and face-shape information. These findings are consistent with color-template approaches used in some computer-based face detection systems.     

Early maturity of face recognition: No childhood development of holistic processing, novel face encoding, or face-space

Historically, it was believed the perceptual mechanisms involved in individuating faces developed only very slowly over the course of childhood, and that adult levels of expertise were not reached until well into adolescence. Over the last 10 years, there has been some erosion of this view by demonstrations that all adult-like behavioural properties are qualitatively present in young children and infants. Determining the age of maturity, however, requires quantitative comparison across age groups, a task made difficult by the need to disentangle development in face perception from development in all the other cognitive factors that affect task performance. Here, we argue that full quantitative maturity is reached early, by 5-7 years at the latest and possibly earlier. This is based on a comprehensive literature review of results in the 5-years-to-adult age range, with particular focus on the results of the few previous studies that are methodologically suitable for quantitative comparison of face effects across age, plus three new experiments testing development of holistic/configural processing (faces versus objects, disproportionate inversion effect), ability to encode novel faces (assessed via implicit memory) and face-space (own-age bias).     

Impaired face recognition does not preclude intact whole face perception

We studied intact and impaired processes in a prosopagnosic patient (RP). In Experiment 1, RP showed an inversion superiority effect with both faces and objects, with better performance when stimuli were presented upside down than in normal upright orientation. In Experiment 2, we studied the effect of face configuration directly by comparing matching performance with normal vs. scrambled faces. RP was worse with normal than with scrambled faces, whereas normal controls showed an advantage of a good face context. In Experiment 3, RP showed interference from external face features on the evaluation of internal face features. These results indicate, first, that although RP is impaired in face recognition and face matching, he does still encode the whole face rather than relying completely on parts-based procedures. Second, RP has a deficit at the level of the configural processes involved in finding subtle differences between individual faces, as his performance is worse when presented with a normal face configuration than with scrambled or inverted faces.     

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.