Introduction to Forensic Face Matching

The role of inversion and face masks on simultaneous and delayed face matching tasks

Although it is generally accepted that face recognition relies on holistic processing, it has been suggested that the simultaneous face matching task may depend on a more analytical or featural processing approach. However, empirical evidence supporting this claim is limited. In two experiments, we further explored the role of holistic and featural processing on simultaneous face matching by manipulating holistic processing through inversion and presenting faces with or without face masks. The results from Experiment 1 revealed that both inversion and face masks impaired matching performance. However, while the inversion effect was evident in both full-view and masked faces, the mask effect was only found in upright, but not inverted, faces. These results were replicated in Experiment 2 but, the inversion and mask effects were stronger in delayed face matching than in simultaneous face matching. Our findings suggest that simultaneous face matching relies on holistic processing, but to a smaller extent compared to higher memory-demanding identification tasks.

High-frequency transcranial random noise stimulation enhances unfamiliar face matching of high resolution and pixelated faces

Face identification is useful for social interactions and its impairment can lead to severe social and mental problems. This ability is also remarkably important in applied settings, including eyewitness identification and ID verification. Several studies have demonstrated the potential of Transcranial Random Noise Stimulation (tRNS) to enhance different cognitive skills. However, research has produced inconclusive results about the effectiveness of tRNS to improve face identification. The present study aims to further explore the effect of tRNS on face identification using an unfamiliar face matching task. Observers firstly received either high-frequency bilateral tRNS or sham stimulation for 20 min. The stimulation targeted occipitotemporal areas, which have been previously involved in face processing. In a subsequent stage, observers were asked to perform an unfamiliar face matching task consisting of unaltered and pixelated face pictures. Compared to the sham stimulation group, the high-frequency tRNS group showed better unfamiliar face matching performance with both unaltered and pixelated faces. Our results show that a single high-frequency tRNS session might suffice to improve face identification abilities. These results have important consequences for the treatment of face recognition disorders, and potential applications in those scenarios whereby the identification of faces is primordial.

Two face masks are better than one: congruency effects in face matching

Although the positive effects of congruency between stimuli are well replicated in face memory paradigms, mixed findings have been found in face matching. Due to the current COVID-19 pandemic, face masks are now very common during daily life outdoor activities. Thus, the present study aims to further explore congruency effects in matching faces partially occluded by surgical masks. Observers performed a face matching task consisting of pairs of faces presented in full view (i.e., full-view condition), pairs of faces in which only one of the faces had a mask (i.e., one-mask condition), and pairs of faces in which both faces had a mask (i.e., two-mask condition). Although face masks disrupted performance in identity match and identity mismatch trials, in match trials, we found better performance in the two-mask condition compared to the one-mask condition. This finding highlights the importance of congruency between stimuli on face matching when telling faces together.

Facial Identity Verification Robust to Pose Variations and Low Image Resolution: Image Comparison Based on Anatomical Facial Landmarks

Face comparison/face mapping is one of the promising methods in face biometrics which needs relatively little effort compared with face identification. Various factors may be used to verify whether two faces are of the same person, among which facial landmarks are one of the most objective indicators due to the same anatomical definition for every face. This study identified major landmarks from 2D and 3D facial images of the same Korean individuals and calculated the distance between the reciprocal landmarks of two images to examine their acceptable range for identifying an individual to obtain standard values from diverse facial angles and image resolutions. Given that reference images obtained in the real-world could be from various angles and resolutions, this study created a 3D face model from multiple 2D images of different angles, and oriented the 3D model to the angle of the reference image to calculate the distance between reciprocal landmarks. In addition, we used the super-resolution method of artificial intelligence to address the inaccurate assessments that low-quality videos can yield. A portion of the process was automated for speed and convenience of face analysis. We conclude that the results of this study could provide a standard for future studies regarding face-to-face analysis to determine if different images are of the same person.

Person Identification from Drones by Humans: Insights from Cognitive Psychology

The deployment of unmanned aerial vehicles (i.e., drones) in military and police operations implies that drones can provide footage that is of sufficient quality to enable the recognition of strategic targets, criminal suspects, and missing persons. On the contrary, evidence from Cognitive Psychology suggests that such identity judgements by humans are already difficult under ideal conditions, and are even more challenging with drone surveillance footage. In this review, we outline the psychological literature on person identification for readers who are interested in the real-world application of drones. We specifically focus on factors that are likely to affect identification performance from drone-recorded footage, such as image quality, and additional person-related information from the body and gait. Based on this work, we suggest that person identification from drones is likely to be very challenging indeed, and that performance in laboratory settings is still very likely to underestimate the difficulty of this task in real-world settings.

A Preliminary Study of the Reliability of Anatomical Facial Landmarks Used in Facial Comparison

Anatomical landmarks are considered the most objective indicators for use in forensic facial comparisons. Therefore, accurately identifying and locating these landmarks is the beginning of reliable facial comparison. This study evaluated the accuracy with which facial landmarks are located and examined their reliability according to type of landmark, head posture, and image quality. Nine operators located a series of landmarks on prepared facial images used to produce comparison images. Then, the average distances between the reciprocal landmarks (ADRL) on the reference and the comparison images were measured as indicators of landmark reliability. We found that a set of landmarks had higher or lower reliability as a function of the head angle and image quality. More reliable landmarks were associated with certain head postures and degrees of image quality. These should be used for facial comparison analysis depending on various head and image conditions.

Species identification by experts and non-experts: comparing images from field guides

Accurate species identification is fundamental when recording ecological data. However, the ability to correctly identify organisms visually is rarely questioned. We investigated how experts and non-experts compared in the identification of bumblebees, a group of insects of considerable conservation concern. Experts and non-experts were asked whether two concurrent bumblebee images depicted the same or two different species. Overall accuracy was below 60% and comparable for experts and non-experts. However, experts were more consistent in their answers when the same images were repeated, and more cautious in committing to a definitive answer. Our findings demonstrate the difficulty of correctly identifying bumblebees using images from field guides. Such error rates need to be accounted for when interpreting species data, whether or not they have been collected by experts. We suggest that investigation of how experts and non-experts make observations should be incorporated into study design, and could be used to improve training in species identification.

Face learning with multiple images leads to fast acquisition of familiarity for specific individuals

Matching unfamiliar faces is a difficult task. Here we ask whether it is possible to improve performance by providing multiple images to support matching. In two experiments we observe that accuracy improves as viewers are provided with additional images on which to base their match. This technique leads to fast learning of an individual, but the effect is identity-specific: Despite large improvements in viewers’ ability to match a particular person's face, these improvements do not generalize to other faces. Experiment 2 demonstrated that trial-by-trial feedback provided no additional benefits over the provision of multiple images. We discuss these results in terms of familiar and unfamiliar face processing and draw out some implications for training regimes.

Factors affecting the identification of individual mountain bongo antelope

The recognition of individuals forms the basis of many endangered species monitoring protocols. This process typically relies on manual recognition techniques. This study aimed to calculate a measure of the error rates inherent within the manual technique and also sought to identify visual traits that aid identification, using the critically endangered mountain bongo, Tragelaphus eurycerus isaaci, as a model system. Identification accuracy was assessed with a matching task that required same/different decisions to side-by-side pairings of individual bongos. Error rates were lowest when only the flanks of bongos were shown, suggesting that the inclusion of other visual traits confounded accuracy. Accuracy was also higher for photographs of captive animals than camera-trap images, and in observers experienced in working with mountain bongos, than those unfamiliar with the sub-species. These results suggest that the removal of non-essential morphological traits from photographs of bongos, the use of high-quality images, and relevant expertise all help increase identification accuracy. Finally, given the rise in automated identification and the use of citizen science, something our results would suggest is applicable within the context of the mountain bongo, this study provides a framework for assessing their accuracy in individual as well as species identification.

Face matching in a long task: enforced rest and desk-switching cannot maintain identification accuracy

In face matching, observers have to decide whether two photographs depict the same person or different people. This task is not only remarkably difficult but accuracy declines further during prolonged testing. The current study investigated whether this decline in long tasks can be eliminated with regular rest-breaks (Experiment 1) or room-switching (Experiment 2). Both experiments replicated the accuracy decline for long face-matching tasks and showed that this could not be eliminated with rest or room-switching. These findings suggest that person identification in applied settings, such as passport control, might be particularly error-prone due to the long and repetitive nature of the task. The experiments also show that it is difficult to counteract these problems.

The shape of the face template: geometric distortions of faces and their detection in natural scenes

Human face detection might be driven by skin-coloured face-shaped templates. To explore this idea, this study compared the detection of faces for which the natural height-to-width ratios were preserved with distorted faces that were stretched vertically or horizontally. The impact of stretching on detection performance was not obvious when faces were equated to their unstretched counterparts in terms of their height or width dimension (Experiment 1). However, stretching impaired detection when the original and distorted faces were matched for their surface area (Experiment 2), and this was found with both vertically and horizontally stretched faces (Experiment 3). This effect was evident in accuracy, response times, and also observers' eye movements to faces. These findings demonstrate that height-to-width ratios are an important component of the cognitive template for face detection. The results also highlight important differences between face detection and face recognition.

Generalization across view in face memory and face matching

While a change in view is considered to be one of the most damaging manipulations for facial identification, this phenomenon has been measured traditionally with tasks that confound perceptual processes with recognition memory. This study explored facial identification with a pairwise matching task to determine whether view generalization is possible when memory factors are minimised. Experiment 1 showed that the detrimental view effect in recognition memory is attenuated in face matching. Moreover, analysis of individual differences revealed that some observers can identify faces across view with perfect accuracy. This was replicated in Experiment 2, which also showed that view generalization is unaffected when only the internal facial features are shown. These results indicate that the view effect in recognition memory does not arise from data limits, whereby faces contain insufficient visual information to allow identification across views. Instead, these findings point to resource limits, within observers, that hamper such person identification in recognition memory.