Morphed Faces Research Articles

Own-species bias in the representations of monkey and human face categories in the primate temporal lobe

Face categorization is fundamental for social interactions of primates and is crucial for determining conspecific groups and mate choice. Current evidence suggests that faces are processed by a set of well-defined brain areas. What is the fine structure of this representation, and how is it affected by visual experience? Here, we investigated the neural representations of human and monkey face categories using realistic three-dimensional morphed faces that spanned the continuum between the two species. We found an "own-species" bias in the categorical representation of human and monkey faces in the monkey inferior temporal cortex at the level of single neurons as well as in the population response analyzed using a pattern classifier. For monkey and human subjects, we also found consistent psychophysical evidence indicative of an own-species bias in face perception. For both behavioural and neural data, the species boundary was shifted away from the center of the morph continuum, for each species toward their own face category. This shift may reflect visual expertise for members of one's own species and be a signature of greater brain resources assigned to the processing of privileged categories. Such boundary shifts may thus serve as sensitive and robust indicators of encoding strength for categories of interest.

Perceptual awareness and categorical representation of faces: Evidence from masked priming

How internal categories influence how we perceive the world is a fundamental question in cognitive sciences. Yet, the relation between perceptual awareness and perceptual categorization has remained largely uncovered so far. Here, we addressed this question by focusing on face perception during subliminal and conscious perception. We used morphed continua between two face identities and we assessed, through a masked priming paradigm, the perceptual processing of these morphed faces under subliminal and supraliminal conditions. We found that priming from subliminal faces followed linearly the information present in the primes, while priming from visible faces revealed a non-linear profile, indicating a categorical processing of face identities. Our results thus point to a special relation between perceptual awareness and categorical processing of faces, and support the dissociation between two modes of information processing: a subliminal mode involving analog treatment of stimuli information, and a supraliminal mode relying on discrete representation.

Emotional sensitivity in youth with borderline personality pathology

If Borderline Personality Disorder (BPD) is characterized by an underlying emotional sensitivity, individuals with this disorder would be expected to demonstrate accurate identification of emotional expressions at earlier stages of expression (i.e., lower thresholds of facial expressivity across all emotional valences). Twenty-one outpatient youth (aged 15–24 years) meeting 3 or more DSM-IV BPD criteria and 20 community-derived participants (aged 15–24 years) with no history of psychiatric problems were tested on a measure of emotional sensitivity, the Face Morph Task. In this test faces morph from neutral to each of the six basic emotional expressions. The BPD group showed no evidence of heightened sensitivity to emotional facial expressions compared to the community control group (all P > 0.05 and effect sizes ranging from 0 to 0.6). They require comparable levels of emotional expressivity in order to correctly identify emotions. Therefore, emotional sensitivity might not be apparent early in the course of BPD. Rather, it might develop later in the course of the disorder or be present only in severe BPD.

3D 스캔 데이터를 이용한 얼굴 애니메이션 시스템

In this paper, we describe the development of a system for generating a 3-dimensional human face using 3D scanned facial data and photo images, and morphing animation. The system comprises a facial feature input tool, a 3-dimensional texture mapping interface, and a 3-dimensional facial morphing interface. The facial feature input tool supports texture mapping and morphing animation - facial morphing areas between two facial models are defined by inputting facial feature points interactively. The texture mapping is done first by means of three photo images - a front and two side images - of a face model. The morphing interface allows for the generation of a morphing animation between corresponding areas of two facial models after texture mapping. This system allows users to interactively generate morphing animations between two facial models, without programming, using 3D scanned facial data and photo images.

Women smokers’ experiences of an age-appearance anti-smoking intervention: A qualitative study

This study was designed to investigate women's experiences of engaging in an age-appearance anti-smoking intervention. Ten 18- to 34-year-old women gave accounts of their experiences after engaging in an age-appearance facial morphing anti-smoking intervention in interviews (n= 7) and a focus group (n= 3), and 37 women gave their accounts while they were engaged in the intervention. Transcripts were analysed using a thematic analysis broadly informed by the procedures of Grounded Theory. Women were very concerned about the impact of ageing on their faces in general, and in particular the additional impact of smoking on their skin. Women were concerned about other people's reactions to them as older smokers with wrinkled skin, and many experienced a physical shock reaction (including reports of nausea) to seeing how they would age if they continued to smoke. They reported that seeing their own face aged on the computer screen increased their perceived risk of skin wrinkling. Women reported being highly motivated to quit smoking as a result of the intervention, and many reported that they would take active steps to quit having seen how they would look if they continued to smoke. This was linked with increased perceived personal responsibility for quitting. Results are discussed in relation to suggestions for anti-smoking interventions aimed at women in the 18- to 34-year-old age group. It is concluded that interventions incorporating age-appearance morphing techniques are likely to be effective in helping women to take active steps to quit smoking.

Manipulating Shape Cues in Dynamic Human Faces: Sexual Dimorphism is Preferred in Female but not Male Faces

AbstractSexual dimorphism in the human face has been linked to attractiveness, and computer‐graphic techniques have been useful in this field by allowing experimental manipulation of dimorphism. However, a limitation of much research is its reliance on static pictorial stimuli, whereas real faces are dynamic, as is much courtship behaviour throughout nature. Furthermore, little is known about possible interactions between static and dynamic facial cues and attractiveness. We adapted well‐established face‐morphing technology to manipulate sexual dimorphism in male and female dynamic facial displays depicting prosocial and antisocial behaviour. Masculinised and feminised versions of these videos were presented as a two‐alternative forced choice task to measure preferences. Feminised female videos were preferred in both movement contexts, as expected. More surprisingly, no directional preference for masculinity or femininity was evident for the male videos in either context. Further analysis showed that the movement of the faces (prosocial vs. antisocial) had no effect on attractiveness ratings. Results were the same with static stimuli and did not conflict with findings from the wider literature using static faces. These findings suggest that the new technique we describe is a valid way to manipulate facial shape in videos and can now be widely applied to future studies of facial morphing.

The contribution of the upper and lower face in happy and sad facial expression classification

We used a happy/sad classification task and a psychophysical model to study the tuning properties of facial expression processors across viewing conditions. Using morphed faces, in this study we measured the extent to which classification of facial expressions depends on the intensity of a particular expression on either the upper or lower face. In the fovea, the upper and lower parts of the test image were either aligned or had a lateral shift of 44′ visual angle. In the periphery, the aligned test image was placed at a 6° visual angle to the left of the fixation. Observers were asked to classify a test image of a facial expression as happy or sad. We discovered that the alignment of the upper and lower halves of the face had no effect on happy/sad classification in the fovea, suggesting that the classification of facial expressions is an analytic process. The model also showed no interaction between the two halves of the face in foveal facial expression classification. In addition, the poor performance of observers in recognizing happiness in the periphery manifests a computational complexity, suggesting a model in which the happy-face processor relies on both facial features and the interaction between them to recognize happiness in the periphery.

Knowing what to look for: Voice affects face race judgements

The study examined the effect that auditory information (speaker language/accent: Japanese or French) had on the processing of visual information (the speaker's race: Asian or Caucasian) in two forced-choice tasks: Classification and perceptual judgement on animated talking characters. Two (male and female) sets of facial morphs were constructed such that a 3-D head of Caucasian appearance was gradually morphed (in 11 steps) into one of Asian appearance. Each facial morph was animated in association with spoken French/Japanese or English with a French/Japanese accent. To examine the auditory effect, each animation was played with or without sound. Experiment 1 used an Asian or Caucasian classification task. Results showed that faces heard in conjunction with Japanese or a Japanese accent were more likely to be classified as Asian compared to those presented without sound. Experiment 2 used a same or different judgement task. Results showed that accuracy was improved by hearing a Japanese accent compared to without sound. These results were discussed in terms of the voice information acting as a cue to assist in organizing and attending to face features.

Flexible kernel memory.

This paper introduces a new model of associative memory, capable of both binary and continuous-valued inputs. Based on kernel theory, the memory model is on one hand a generalization of Radial Basis Function networks and, on the other, is in feature space, analogous to a Hopfield network. Attractors can be added, deleted, and updated on-line simply, without harming existing memories, and the number of attractors is independent of input dimension. Input vectors do not have to adhere to a fixed or bounded dimensionality; they can increase and decrease it without relearning previous memories. A memory consolidation process enables the network to generalize concepts and form clusters of input data, which outperforms many unsupervised clustering techniques; this process is demonstrated on handwritten digits from MNIST. Another process, reminiscent of memory reconsolidation is introduced, in which existing memories are refreshed and tuned with new inputs; this process is demonstrated on series of morphed faces.

Scanning strategies do not modulate face identification: eye-tracking and near-infrared spectroscopy study.

BackgroundDuring face identification in humans, facial information is sampled (seeing) and handled (processing) in ways that are influenced by the kind of facial image type, such as a self-image or an image of another face. However, the relationship between seeing and information processing is seldom considered. In this study, we aimed to reveal this relationship using simultaneous eye-tracking measurements and near-infrared spectroscopy (NIRS) in face identification tasks.Methodology/Principal Findings22 healthy adult subjects (8 males and 14 females) were shown facial morphing movies in which an initial facial image gradually changed into another facial image (that is, the subject's own face was changed to a familiar face). The fixation patterns on facial features were recorded, along with changes in oxyhemoglobin (oxyHb) levels in the frontal lobe, while the subjects identified several faces. In the self-face condition (self-face as the initial image), hemodynamic activity around the right inferior frontal gyrus (IFG) was significantly greater than in the familiar-face condition. On the other hand, the scanning strategy was similar in almost all conditions with more fixations on the eyes and nose than on other areas. Fixation time on the eye area did not correlate with changes in oxyHb levels, and none of the scanning strategy indices could estimate the hemodynamic changes.Conclusions/SignificanceWe conclude that hemodynamic activity, i.e., the means of processing facial information, is not always modulated by the face-scanning strategy, i.e., the way of seeing, and that the right IFG plays important roles in both self-other facial discrimination and self-evaluation.

Categorical Perception for Unfamiliar Faces

Robust findings show that categorical perception (CP) occurs in identification of familiar faces. CP has also been observed for unfamiliar morphed faces after sufficient learning of the original, unmorphed faces has taken place. We previously suggested that CP arises when the activation of inconsistent visual and verbal representations creates a conflict between perceptual and category information. In the present study, we conducted two experiments in which the endpoint faces of an unfamiliar morphed continuum were presented in either a covert training regime (famous vs. nonfamous judgments) or an overt training regime (previously seen vs. unseen judgments). In both experiments, participants’ reaction times to repeated targets decreased relative to reaction times to control items during training. After overt training, CP was observed for the previously unfamiliar faces. No CP was observed for covertly trained faces. We conclude that individual faces must be explicitly categorized before CP can be established for the morphed continuum between them.

Morphed emotional faces: Emotion detection and misinterpretation in social anxiety

The current study investigated detection and interpretation of emotional facial expressions in high socially anxious (HSA) individuals compared to non-anxious controls (NAC). A version of the morphed faces task was implemented to assess emotion onset perception, decoding accuracy and interpretation, either with time pressure (Restricted Viewing Task, RVT) or with unlimited viewing (Free Viewing Task, FVT). Twenty-seven HSA and 30 NAC viewed sequences of neutral faces slowly changing to full-intensity angry, happy, or disgust expressions. Participants were instructed to assign the expression as soon as possible to one of four given emotion categories (angry, contempt, disgust, or happy). While no group differences were found for emotion onset perception or decoding performance, the results suggest an interpretation bias in HSA. Under the RVT condition, HSA demonstrated a threat bias (disgust interpreted as contempt), contrasting the NAC’s positive bias (disgust interpreted as happy). No group differences were found in the FVT. We suggest that socially anxious individuals tend to misinterpret facial expressions as threatening when they must do so quickly and efficiently, as in real life.

Acquired prosopagnosia as a face-specific disorder: Ruling out the general visual similarity account

Prosopagnosia is classically defined as a disorder of visual recognition specific to faces, following brain damage. However, according to a long-standing alternative view, these patients would rather be generally impaired in recognizing objects belonging to visually homogenous categories, including faces. We tested this alternative hypothesis stringently with a well-documented brain-damaged prosopagnosic patient (PS) in three delayed forced-choice recognition experiments in which visual similarity between a target and its distractor was manipulated parametrically: novel 3D geometric shapes, morphed pictures of common objects, and morphed photographs of a highly homogenous familiar category (cars). In all experiments, PS showed normal performance and speed, and there was no evidence of a steeper increase of error rates and RTs with increasing levels of visual similarity, compared to controls. These data rule out an account of acquired prosopagnosia in terms of a more general impairment in recognizing objects from visually homogenous categories. An additional experiment with morphed faces confirmed that PS was specifically impaired at individual face recognition. However, in stark contrast to the alternative view of prosopagnosia, PS was relatively more impaired at the easiest levels of discrimination, i.e. when individual faces differ clearly in global shape rather than when faces were highly similar and had to be discriminated based on fine-grained details. Overall, these observations as well as a review of previous evidence, lead us to conclude that this alternative view of prosopagnosia does not hold. Rather, it seems that brain damage in adulthood may lead to selective recognition impairment for faces, perhaps the only category of visual stimuli for which holistic/configural perception is not only potentially at play, but is strictly necessary to individualize members of the category efficiently.

Attentional modulation of face expression perception

Previous studies in our lab have shown that the attentional state (attending versus ignoring) that is applied when initially exposed to a novel face can determine later how that face will be evaluated on social emotional scales. Specifically, faces that are ignored in difficult attention tasks are later seen as less trustworthy than attended faces. A possible explanation for attentional modulation of affective appraisal of faces is that prior active ignoring subtly alters subsequent sensitivity to face expression. To test this possibility, we made careful psychometric measurements of face expression perception using face morphs (e.g., angry-neutral; happy-neutral) of the same individuals expressing different expressions. Participants judged faces with a range of morph values as neutral versus expressive. The resultant psychometric curves (% “expressive” response plotted as a function of % face expression intensity) were used to select approximately equivalently intense expressions for each face and expression dimension in the stimulus set. Faces with these values were then used in the main experiment. Here, observers engaged in a simple but demanding visual search task (to control their attention state) using neutral faces only. Immediately after each search trial, they made by a face expression judgement (using the morph values derived from our psychometric measurements) on a face that was either a prior target or distractor. The post-search psychometric functions were then used to reveal whether prior attention modulated the slope or shifted the expression sensitivity functions laterally. The former indicates sensitivity change whereas the latter reflects a biasing mechanism. Results for both happy and angry expression will be discussed in terms of attentional modulation of perception in other stimulus domains.

Holistic face encoding is modulated by perceived face race: Evidence from perceptual adaptation

Human expertise at processing faces relies on how facial features are encoded: as a whole template rather than as a sum of independent features. This holistic encoding is less prominent for other-race faces, possibly accounting for the difficulty one encounters in recognizing these faces (the ‘other-race effect’). Here, we tested the hypothesis that the magnitude of holistic face encoding can be modulated by racial categorization of the face. Caucasian participants performed a face-composite task with ‘racially-ambiguous’ face-stimuli (cross-race morphed faces, equally categorized as Asian or Caucasian faces in an independent task). The perceived race of the ambiguous faces was manipulated using adaptation. Experiment 1 showed that identical morphed face-stimuli were processed more holistically when perceived as ‘same-race’ than as ‘other-race’, i.e., following adaptation to ‘other-race’ versus ‘same-race’, respectively. Experiment 2 ascertained that the determining factor in the observed holistic processing modulation was the race of the racially-ambiguous face as perceived, rather than expected, by the participants, which supports the idea that the holistic processing of the face-stimuli was modulated by their race-categorization at the perceptual level.

Reduced neural selectivity increases fMRI adaptation with age during face discrimination

Ventral–visual activity in older adults has been characterized by dedifferentiation, or reduced distinctiveness, of responses to different categories of visual stimuli such as faces and houses, that typically elicit highly specialized responses in the fusiform and parahippocampal brain regions respectively in young adults (Park et al., 2004). In the present study, we demonstrate that age-related neural dedifferentiation applies to within-category stimuli (different types of faces) as well, such that older adults process less distinctive representations for individual faces than young adults. We performed a functional magnetic resonance imaging adaptation experiment while young and older participants made same-different judgments to serially presented face-pairs that were Identical, Moderate in similarity through morphing, or Different. As expected, older adults showed adaptation in the fusiform face area (FFA), during the Identical as well as the Moderate conditions relative to the Different condition. Young adults showed adaptation during the Identical condition, but minimal adaptation to the Moderate condition. These results indicate that older adults' FFA treated the morphed faces as Identical faces, reflecting decreased fidelity of neural representation of faces with age.

Does perceived race affect discrimination and recognition of ambiguous-race faces? A test of the sociocognitive hypothesis.

Discrimination and recognition are often poorer for other-race than own-race faces. These other-race effects (OREs) have traditionally been attributed to reduced perceptual expertise, resulting from more limited experience, with other-race faces. However, recent findings suggest that sociocognitive factors, such as reduced motivation to individuate other-race faces, may also contribute. If the sociocognitive hypothesis is correct, then it should be possible to alter discrimination and memory performance for identical faces by altering their perceived race. We made identical ambiguous-race morphed faces look either Asian or Caucasian by presenting them in Caucasian or Asian face contexts, respectively. However, this perceived-race manipulation had no effect on either discrimination (Experiment 1) or memory (Experiment 2) for the ambiguous-race faces, despite the presence of the usual OREs in discrimination and recognition of unambiguous Asian and Caucasian faces in our participant population. These results provide no support for the sociocognitive hypothesis. (PsycINFO Database Record (c) 2009 APA, all rights reserved).

Prediction of human eye movements in facial discrimination tasks

Under natural viewing conditions, human observers selectively allocate their attention to subsets of the visual input. Since overt allocation of attention appears as eye movements, the mechanism of selective attention can be uncovered through computational studies of eyemovement predictions. Since top-down attentional control in a task is expected to modulate eye movements significantly, the models that take a bottom-up approach based on low-level local properties are not expected to suffice for prediction. In this study, we introduce two representative models, apply them to a facial discrimination task with morphed face images, and evaluate their performance by comparing them with the human eye-movement data. The result shows that they are not good at predicting eye movements in this task.

Categorical Perception of emotional faces is not affected by aging

Effects of normal aging on categorical perception (CP) of facial emotional expressions were investigated. One-hundred healthy participants (20 to 70 years old; five age groups) had to identify morphed expressions ranging from neutrality to happiness, sadness and fear. We analysed percentages and latencies of correct recognition for nonmorphed emotional expressions, percentages and latencies of emotional recognition for morphed-faces, locus of the boundaries along the different continua and the number of intrusions. The results showed that unmorphed happy and fearful faces were better processed than unmorphed sad and neutral faces. For morphed faces, CP was confirmed, as latencies increased as a function of the distance between the displayed morph and the original unmorphed photograph. The locus of categorical boundaries was not affected by age. Aging did not alter the accuracy of recognition for original pictures, no more than the emotional recognition of morphed faces or the rate of intrusions. However, latencies of responses increased with age, for both unmorphed and morphed pictures. In conclusion, CP of facial expressions appears to be spared in aging.

The Secret Life of Kernels: Reconsolidation in Flexible memories

Event Abstract Back to Event The Secret Life of Kernels: Reconsolidation in Flexible memories In this paper we are presenting memory reconsolidation in kernel associative memories. Reconsolidation is an important process in memory dynamics that is observed both in neurophysiological and psychological studies, as well as modelled in various artificial neural systems. The memory tracks changes in the environment or in associations among objects. Recent models predict that memory representations should be sensitive to learning order, consistent with psychophysical studies of face recognition and electrophysiological experiments on hippocampal place cells. Our goal is to show that such reconsolidation effects are possible in more flexible environment, dealing with large-scale data. As memory model we introduce a special neural system that while relying on some of the analysis developed by Hopfield has memory attractors that do not lie in the input or neural space but rather in an abstract unbounded high-level kernel space. For reconsolidation we choose the principle of global memory update. It is more stable than updating the attractor closest to current input, and also this technique enables direct analogy to existing reconsolidation methods in classical Hopfield networks. We establish a metric in between kernel associative memories. In the feature space this metric is equivalent to the distance between the weight matrices of two networks, but in input space it is a Riemannian metric. Subsequent application of this procedure implies dynamic reconsolidation of memories that stay always consistent to the changing environment. Basing on this metric we construct an elementary update procedure - one step of reconsolidation. Memory is shifted in the direction of new input along a geodesic curve in corresponding Riemannian space. Our first experiment was made using the MNIST database of handwritten digits. We examined network's ability to track images gradually varying in time. For this purpose a learning set of rotating digits was created. The learning set contained 9000 images. They were obtained from 100 original digits (10 per class) by rotating them counterclockwise on angle from 0 to 180o.Classification was tested on the set of 1000 images closest to the final state. Obtained recognition quality was 96.4%. The second experiment is similar to one previously made in humans. Attractor tracking was investigated using sequences of morphed faces from Productive Aging Lab's Face Database. When the learning order follows image order in the morphing sequence, attractors changed gradually and consistently. The ability to recognize the initial set of images gradually decreased when attractors tended to the final set. In case of random learning order attractors quickly got messy, with no significant ability to distinguish faces. This experiment also demonstrates efficiency of reconsolidation in kernel memories for high-dimensional data. Based on these results we conclude that reconsolidation in kernel memories is both computationally efficient and biologically plausible, and it can model phenomena previously observed in human memory. Conference: Computational and systems neuroscience 2009, Salt Lake City, UT, United States, 26 Feb - 3 Mar, 2009. Presentation Type: Poster Presentation Topic: Poster Presentations Citation: (2009). The Secret Life of Kernels: Reconsolidation in Flexible memories. Front. Syst. Neurosci. Conference Abstract: Computational and systems neuroscience 2009. doi: 10.3389/conf.neuro.06.2009.03.271 Copyright: The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers. They are made available through the Frontiers publishing platform as a service to conference organizers and presenters. The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated. Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed. For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions. Received: 04 Feb 2009; Published Online: 04 Feb 2009. Login Required This action requires you to be registered with Frontiers and logged in. To register or login click here. Abstract Info Abstract Supplemental Data The Authors in Frontiers Google Google Scholar PubMed Related Article in Frontiers Google Scholar PubMed Abstract Close Back to top Javascript is disabled. Please enable Javascript in your browser settings in order to see all the content on this page.