Face perception and recognition

Face makeup is applied to cover unwanted marked on the face in order to improve a person's appearance. However, this is used in criminal activities since makeup can disguise the true identity of the person. This concern serves as the basis by several studies to discover the technique and applied different methods to optimize the detection and improve accuracy. This study aims to conduct an experiment by applying new methods and techniques in order to provide new results which will increase the scope to consider. Furthermore, this study utilized best performing algorithms: the Haar-Cascade classifier for face detection, Color Feature Extraction for detection of face makeup and Local Binary Pattern Histogram algorithm for extraction and recognition of the features which applied in the experiments considering condition from fluorescent light and sunlight, angle variation of 0 degree, 45% upward and downward and distance of 0.5 meter and 1 meter. The study collected an overall of 3000 images which will serve as the training datasets. In result, the experiment marked an average accuracy of 88.75% comprises all the methods applied. Specifically, the result showed an average result of both sunlight and fluorescent result with no makeup marked an average score of 90% while the result with makeup marked a score of 87.5%. Moreover, the results showed a higher recognition in near distance of 0.5 meters, the fluorescent light recorded an accuracy result of 90% and 0% degree marked a score of 100%.

Smiles are distinct and easily recognizable facial expressions, yet they markedly differ in their meanings. According to a recent theoretical account, smiles can be classified based on three fundamental social functions which they serve: expressing positive affect and rewarding self and others (reward smile), creating and maintaining social bonds (affiliative smile), and negotiating social status (dominance smiles) (Niedenthal et al., 2010; Martin et al., 2017). While there is evidence for distinct morphological features of these smiles, their categorization only starts to be investigated in human faces. Moreover, the factors influencing this process – such as facial mimicry or display mode – remain yet unknown. In the present study, we examine the recognition of reward, affiliative, and dominance smiles in static and dynamic portrayals, and explore how interfering with facial mimicry affects such classification. Participants (N = 190) were presented with either static or dynamic displays of the three smile types, whilst their ability to mimic was free or restricted via a pen-in-mouth procedure. For each stimulus they rated the extent to which the expression represents a reward, an affiliative, or a dominance smile. Higher than chance accuracy rates revealed that participants were generally able to differentiate between the three smile types. In line with our predictions, recognition performance was lower in the static than dynamic condition, but this difference was only significant for affiliative smiles. No significant effects of facial muscle restriction were observed, suggesting that the ability to mimic might not be necessary for the distinction between the three functional smiles. Together, our findings support previous evidence on reward, affiliative, and dominance smiles by documenting their perceptual distinctiveness. They also replicate extant observations on the dynamic advantage in expression perception and suggest that this effect may be especially pronounced in the case of ambiguous facial expressions, such as affiliative smiles.

Applied behavioral analysis (ABA) is an effective form of therapy for children with autism spectrum disorder (ASD), but it faces criticism for being un-generalizable, too time intensive, and too dependent on specialists to deliver treatment. Earlier age at onset of therapy is one of the strongest predictors of later success, but waitlists to begin therapies can be as long as 18 months. To combat complications associated with the clinical setting and expedite access to therapy, we have begun development of Autism Glass, a machine-learning-assisted software system that runs on Google Glass and an Android Smartphone; it is designed for use in the child’s natural environment during social interactions.This is an exploratory- and co-designed-based study to see how children with ASD respond to our device and examine preliminary data on its effectiveness.

FACES 2.0 is a face recognition application based on a method of machine learning known as deep neural networking.  Funded by the Kress Foundation (with initial work funded by the National Endowment for the Humanities), it is designed to automatically test the degree of probability of a shared identification between different works of portrait art--that is, non-photographic portraits that are subject to the subjectivity of artistic interpretation.  When used correctly with good images, it has the potential to match what is known with a given unknown--something that is unlikely to be accidental--and yields results that may be considered probable.  In this, FACES has the potential to provide previously unnoticeable or unconfirmable information by contributing categories of quantifiable data for researchers to factor into their own analyses.
The FACES website will be available sometime in January, 2018: http://faces.ucr.edu

Opposing aftereffects can be simultaneously induced by adapting to faces of different races distorted in opposite directions, allowing researchers to infer that faces are encoded against race-specific prototypes. This effect also suggests the existence of dissociable pools of neurons sensitive to race, each of which has been differently adapted to cause an opposite aftereffect. More recent studies have suggested that changes in the strength of race-contingent aftereffects reveal evidence of categorical perception, as they are larger when the adapting faces straddle the racial category boundary. We examined whether changes in these effects more closely correspond to a dichotomous categorical judgment, reflecting highly race-selective neural mechanisms, or more continuous perceptions of racial typicality, reflecting visual channels that are more broadly tuned. In Experiment 1, faces with a range of “morph levels” (i.e., relative contributions of Asian/Caucasian faces) were either rated on a continuous scale for Asian/Caucasian typicality, or simply categorized as Asian/Caucasian. As expected, typicality ratings showed a shallow slope (observers were sensitive to morph level over a broad range), while dichotomous racial categorization showed a steep slope (a rapid switch from categorization as Asian-Caucasian). In Experiment 2, race-contingent adaptation was assessed using test faces with various morph levels. Aftereffect size showed a shallow slope, closely resembling the racial typicality ratings, but showing a significant difference to the categorization data. This suggests that although the visual channels processing these faces do show some selectivity to race, they are sensitive to perceptions of racial typicality, showing a gradual transition of activity across a broad range of faces along the racial continuum.

In the current study we use contingent face aftereffects to examine the contributions made by morphology and facial luminance cues to the encoding of race. African and European facial images differed in both morphology and luminance (ML), in morphology alone (M), or in luminance alone (L). Significant aftereffects were found in conditions where test stimuli included morphological information (ML and M), but not when it was absent (L). Furthermore, the size of the aftereffect for test conditions specifying race using both cues (ML) was no greater than when morphological information was the sole cue (M). This suggests that the effects measured here can be accounted for by differences in morphology alone. These results indicate that not only are individual reports of perceived racial typicality more significantly influenced by morphological shape cues than surface cues, but morphological cues appear to form the basis of the underlying neural encoding for faces of different races.

Recent work demonstrates that human face identification is most efficient when based on horizontal, rather than vertical, image structure. Because it is unclear how this specialization for upright (compared to inverted) face processing emerges in the visual system, the present study aimed to systematically characterize the orientation sensitivity profile for face identification. With upright faces, identification performance in a delayed match-to-sample task was highest for horizontally filtered images and declined sharply with oblique and vertically filtered images. Performance was well described by a Gaussian function with a bandwidth around 25°. Face inversion reshaped this sensitivity profile dramatically, with a downward shift of the entire tuning curve as well as a reduction in the amplitude of the horizontal peak and a doubling in bandwidth. The use of naturalistic outer contours (vs. a common outline mask) was also found to reshape this sensitivity profile by increasing sensitivity to oblique information in the near-horizontal range. Altogether, although face identification is sharply tuned to horizontal angles, both inversion and outline masking can profoundly reshape this orientation sensitivity profile. This combination of image-and observer-driven effects provides an insight into the functional relationship between orientation-selective processes within primary and high-level stages of the human brain. The remarkable efficiency of human face recognition presents a challenge to our understanding of the visual system. Research on the face inversion effect (FIE) indicates that this efficiency is likely derived from observer-driven biases at high-level stages of visual processing that are selective for the higher-order properties of upright faces (e.g., features like eyes, nose, and mouth and their canonical spatial arrangement within the face 1–5). How such high-level specialization emerges from earlier stages of visual encoding is largely unknown. In the brain, visual processing begins with the encoding of local edge orientation in primary visual cortex (V1). The responses of V1 neurons typically reveal a preferred orientation that produces a peak level of response, with decreasing responsiveness for edges angled away from the preferred orientation 6,7. The function relating V1 neuronal responses to orientation is a Gaussian whose bandwidth reflects the sharpness of the tuning to orientation. Electrophysiological recordings in monkey V1 have shown that the tuning bandwidth for most single neurons is between 20 and 50° full-width at half-maximum 8,9. Mechanisms with similar tuning bandwidths have been argued to underlie the human perception of tilt contrast and the tilt aftereffect 10,11. However, it is clear that our efficiency at processing faces does not arise at this computational level. The inversion of a face in the picture plane produces a dramatic impairment in face perception 12 despite the fact that it preserves the orientation content of the image (aside from flipping its sign) and should not therefore significantly alter V1 responses at the population level 13. Recent evidence nonetheless suggests that the high-level mechanisms selectively recruited for upright face identification are orientation-selective. When upright faces are filtered to preserve information in narrow orientation ranges (Fig. 1a), human identification performance peaks in the horizontal band, declines progressively as stimulus orientation departs from horizontal, and reaches its minimum in the vertical range 14. With face inversion, however, human sensitivity to facial identity becomes comparable for horizontal and vertical image structure. These findings suggest that high-level face-specialized mechanisms are particularly sensitive to the horizontal range of facial information 15–17. Accordingly, the removal of the horizontal components of the facial image has been found to disrupt face-specialized processing (as marked by the FIE) in several of the high-level visual regions responding preferentially to faces in the human and monkey brain (i.e., FFA 13,18,19).

Verbal descriptions of unfamiliar faces have been found to impair later identification of these faces in adults, a phenomenon known as the “verbal overshadowing effect” (VOE). Although determining whether children are good at describing unfamiliar individuals and whether these descriptions impair their recognition performance is critical to gaining a better understanding children's eyewitness ability, only a couple of studies have examined this dual issue in children and these found no evidence of VOE. However, as there are some methodological criticisms of these studies, we decided to conduct two further experiments in 7–8, 10–11, and 13–14-year-old children and in adults using a more optimal method for the VOE to be observed. Evidence of the VOE on face identification was found in both children and adults. Moreover, neither the accuracy of descriptions, nor delay nor target presence in the lineup was found to be associated with identification accuracy. The theoretical and developmental implications of these findings are discussed.