Face Masks Do Not Alter Gaze Cueing of Attention: Evidence From the COVID-19 Pandemic

Face masks and fake masks: the effect of real and superimposed masks on face matching with super-recognisers, typical observers, and algorithms

Mask wearing has been required in various settings since the outbreak of COVID-19, and research has shown that identity judgements are difficult for faces wearing masks. To date, however, the majority of experiments on face identification with masked faces tested humans and computer algorithms using images with superimposed masks rather than images of people wearing real face coverings. In three experiments we test humans (control participants and super-recognisers) and algorithms with images showing different types of face coverings. In all experiments we tested matching concealed or unconcealed faces to an unconcealed reference image, and we found a consistent decrease in face matching accuracy with masked compared to unconcealed faces. In Experiment 1, typical human observers were most accurate at face matching with unconcealed images, and poorer for three different types of superimposed mask conditions. In Experiment 2, we tested both typical observers and super-recognisers with superimposed and real face masks, and found that performance was poorer for real compared to superimposed masks. The same pattern was observed in Experiment 3 with algorithms. Our results highlight the importance of testing both humans and algorithms with real face masks, as using only superimposed masks may underestimate their detrimental effect on face identification.

A positive side effect of wearing face coverings for socially anxious females: Findings from a speech task

Background

Wearing face masks has become more common due to the COVID-19 pandemic. Until now, it has not been investigated whether socially anxious individuals can reduce their acute anxiety symptoms by wearing face masks during a speech task.

Method

Fifty-nine socially anxious females were asked to prepare and give an oral presentation. Participants were randomly allocated either to a group that was asked to wear a face mask during the task, or to a group that was not asked to wear a face mask during the task. Dependent variables included physiological parameters (systolic/diastolic blood pressure, heart rate) and self-reports (valence and arousal at baseline, as well as directly before and after the presentation).

Results

The analyses showed that the group without face masks had a higher systolic blood pressure than the group with face masks, directly before as well as after the presentation. The two groups did not differ in the other measures.

Conclusion

Wearing face masks has a small stress-reducing effect on socially anxious females.

Foot cues can elicit covert orienting of attention

Humans tend to orient their attentional resources towards the same location indicated by spatial signals coming from the others, such as pointing fingers, head turns, or eye-gaze. Here, two experiments investigated whether an attentional orienting response can be elicited even by foot cues. Participants were asked to localize a peripheral target while a task-irrelevant picture of a naked human foot, oriented leftward or rightward, was presented on the centre of the screen. The foot appeared in a neutral posture (i.e., standing upright) or an action-oriented posture (i.e., walking/running). In Experiment 1, neutral and action-oriented feet were presented in two distinct blocks, while in Experiment 2 they were presented intermixed. The results showed that the action-oriented foot, but not the neutral one, elicited an orienting response, though this only emerged in Experiment 2. This work suggests that attentional shifts can be induced by action-oriented foot cues, as long as these stimuli are made contextually salient.

Influence of colored face masks on judgments of facial attractiveness and gaze patterns

BACKGROUND

Facial aesthetics are of great importance in social interaction. With the widespread adoption of face masks in response to the Covid-19 pandemic, there is growing interest in understanding how wearing masks might impact perceptions of attractiveness, as they partially or completely conceal facial features that are typically associated with attractiveness.

OBJECTIVES

This study aimed to explore the impact of mask wearing on attractiveness and to investigate whether the color (red or blue) of the mask has any effect on the perception of a person's attractiveness, while also considering gender and age as contributing factors. Additionally, the study intended to evaluate gaze patterns, initial focus, and dwell time in response to masked and unmasked faces.

METHODS

30 AI-generated images of 15 female and 15 male faces were presented to 71 participants (35 male, 36 female) in 3 conditions: not wearing any mask, wearing a red surgical mask, and wearing a blue surgical mask. The perceived attractiveness was rated on an ordinal scale of 1-10 (10 being most attractive). Gaze behavior, dwell time and initial focus were recorded using a stationary eye-tracking system.

RESULTS

The study found that wearing masks had no significant effect on the attractiveness ratings of female faces (p = .084), but it did benefit the perceived attractiveness of male faces which were initially rated lower (p = .16). Gender and age also played a significant role, as both male and female participants rated female stimuli higher than male stimuli (p < .001), and younger participants rated both genders as less attractive than older participants (p < .01). However, there was no significant influence of the mask's color on attractiveness. During the eye-tracking analysis, the periorbital region was of greater interest while masked, with the time to first fixation for the eyes being lower than the non-masked stimulus (p < .001) and showed a longer dwell time (p < .001). The lower face was shown less interest while masked as the time to first fixation was higher (p < .001) and the fixation count was less (p < .001). Mask color did not influence the scan path and there was no difference in revisits to the mask area between red or blue masks (p = .202), nor was there a difference in time to first fixation (p = .660).

CONCLUSIONS

The study findings indicate that there is an interplay between the gender and age of the participant and the facial stimuli. The color red did have an effect on the perception attractiveness, however not in female faces. The results suggest that masks, especially red ones, might be more beneficial for male faces, which were perceived as less attractive without a mask. However, wearing a mask did not significantly impact already attractive faces. The eye-tracking results revealed that the periorbital region attracted more attention and was fixated on more quickly while wearing a mask, indicating the importance of eyes in social interaction and aesthetic perception.

Face mask reduces gaze-cueing effect

Recent studies have found that face masks affect social cognition and behaviour in the context of the novel coronavirus (COVID-19) pandemic. The eyes, the only part of the face not covered by face masks, are an important spatial attention cue that can trigger social attention orienting. Here, we adopted a spatial gaze-cueing task to investigate whether face masks affect social attention orienting triggered by eye gaze cues. In Experiment 1, participants were asked to determine the orientation of a target line under two types of cues-masked and non-masked faces-and two stimulus onset asynchrony (SOA) conditions (300 ms and 1000 ms). The results showed that masked faces induced a smaller gaze-cueing effect (GCE) compared to non-masked faces at 300 ms SOA, while two face types induced similar GCEs at 1000 ms SOA. Experiment 2 used mouth-obscured faces and non-masked faces as cues and found that no significant difference in GCE between the two types at either 300 ms or 1000 ms SOA, indicating that the reduction of GCE caused by the masked face was due to the social meaning expressed by the mask rather than a physical effect of masking. The present study extends previous findings to support the idea that high-level social information affects the processing of eye gaze direction and provides evidence that face masks affect social cognition and behaviour in the context of COVID-19.

A Trained Humanoid Robot can Perform Human-Like Crossmodal Social Attention and Conflict Resolution

To enhance human-robot social interaction, it is essential for robots to process multiple social cues in a complex real-world environment. However, incongruency of input information across modalities is inevitable and could be challenging for robots to process. To tackle this challenge, our study adopted the neurorobotic paradigm of crossmodal conflict resolution to make a robot express human-like social attention. A behavioural experiment was conducted on 37 participants for the human study. We designed a round-table meeting scenario with three animated avatars to improve ecological validity. Each avatar wore a medical mask to obscure the facial cues of the nose, mouth, and jaw. The central avatar shifted its eye gaze while the peripheral avatars generated sound. Gaze direction and sound locations were either spatially congruent or incongruent. We observed that the central avatar's dynamic gaze could trigger crossmodal social attention responses. In particular, human performance was better under the congruent audio-visual condition than the incongruent condition. Our saliency prediction model was trained to detect social cues, predict audio-visual saliency, and attend selectively for the robot study. After mounting the trained model on the iCub, the robot was exposed to laboratory conditions similar to the human experiment. While the human performance was overall superior, our trained model demonstrated that it could replicate attention responses similar to humans.

Guilt-inducing interaction with others modulates subsequent attentional orienting via their gaze

Gaze direction can trigger social attentional orientation, characterised by a speeded reaction time in detecting targets appearing in a gazed-at location compared with those appearing in other locations. This is called the 'gaze-cueing effect' (GCE). Here, we investigated whether a feeling of guilt established through prior interaction with a cueing face could modulate the gaze-cueing effect. Participants first completed a guilt-induction task using a modified dot-estimation paradigm to associate the feeling of guilt with a specific face, after which the face that had established the binding relationship was used as the stimulus in a gaze-cueing task. The results showed that guilt-directed faces and control faces induce equal magnitudes of gaze-cueing effect in 200 ms of stimulus onset asynchrony (SOA), while guilt-directed faces induce a smaller gaze-cueing effect than control faces in 700 ms SOA. These findings provide preliminary evidence that guilt may modulate social attention triggered by eye gaze at a later stage of processing but not in the earlier stages.

Face masks have a limited effect on the feeling of being looked at

Introduction

Wearing face masks has been promoted as an effective measure to reduce the spread of COVID-19. Because face masks cover a major part of the face, they have detrimental effects on various aspects of social cognition. Yet, a highly important feature of the face is not occluded by face masks: the eyes. The eyes play an important role in social interactions: knowing where another person is looking is of central importance when interacting with others. Recent research has reported an attentional shift toward the eye region as a consequence of the widespread exposure to face masks. However, no study has yet investigated the influence of face masks on the perception of eye gaze direction. Here we investigated whether face masks have an effect on the feeling of being looked at. Assuming an attentional shift toward the eyes, we might expect more accurate gaze perception in faces wearing face masks.

Methods

Sixty-five participants decided for a series of realistic avatar faces whether each face was making eye contact or not. Half of the faces wore face masks, the other half did not. For each participant and separately for each condition (mask vs. no mask), we calculated the cone of direct gaze (CoDG), a commonly used measure to quantify the range of gaze angles within which an observer assumes mutual gaze.

Results

Contrary to our expectations, results show that mutual gaze is not recognized more accurately in masked faces. Rather, the CoDG was, on average, slightly wider for faces wearing masks compared to faces without masks.

Discussion

Notwithstanding the relatively small effect of face mask, these findings potentially have implications on our social interactions. If we inadvertently feel looked at by an onlooker, we may react inappropriately by reciprocating the alleged approach orientation.

Face mask use as a categorical dimension in social perception

Prevention measures aimed at combating COVID-19 pandemic strongly impact several aspects of social life. In particular, interpersonal perception is affected as a function of whether the persons perceived wear or not face masks. In two experimental studies, we here explored whether people rely on the presence vs. absence of face masks when encoding information in memory about other individuals. In a memory confusion paradigm, participants were initially presented with individuals either wearing a face mask or not, each conveying a series of sentences. Next, participants were probed about the identity of the speaker of each sentence. Results showed that it was more likely to erroneously attribute a sentence to a speaker who also was wearing a face mask (or not) as the original speaker, demonstrating that the cue about wearing or not a face mask was spontaneously used to encode information. Study 2 ruled out an alternative explanation based on perceptual processes, suggesting that face masks represent meaningful social objects. Overall, it emerged that participants spontaneously categorize others as a function of whether they wear a mask or not. Findings also confirmed previous research evidence about the more positive evaluation of mask wearers as compared to non-wearers, and the overall detrimental impact that face masks may have on the correct identification of social targets.

Face masks inhibit facial cues for approachability and trustworthiness: an eyetracking study

Wearing face masks during the Covid-19 pandemic has undeniable benefits from our health perspective. However, the interpersonal costs on social interactions may have been underappreciated. Because masks obscure critical facial regions signaling approach/avoidance intent and social trust, this implies that facial inference of approachability and trustworthiness may be severely discounted. Here, in our eyetracking experiment, we show that people judged masked faces as less approachable and trustworthy. Further analyses showed that the attention directed towards the eye region relative to the mouth region mediated the effect on approachability, but not on trustworthiness. This is because for masked faces, with the mouth region obscured, visual attention is then automatically diverted away from the mouth and towards the eye region, which is an undiagnostic cue for judging a target's approachability. Together, these findings support that mask-wearing inhibits the critical facial cues needed for social judgements.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12144-022-03705-8.

Wearing the face mask affects our social attention over space

Recent studies suggest that covering the face inhibits the recognition of identity and emotional expressions. However, it might also make the eyes more salient, since they are a reliable index to orient our social and spatial attention. This study investigates (1) whether the pervasive interaction with people with face masks fostered by the COVID-19 pandemic modulates the processing of spatial information essential to shift attention according to other’s eye-gaze direction (i.e., gaze-cueing effect: GCE), and (2) whether this potential modulation interacts with motor responses (i.e., Simon effect). Participants were presented with face cues orienting their gaze to a congruent or incongruent target letter location (gaze-cueing paradigm) while wearing a surgical mask (Mask), a patch (Control), or nothing (No-Mask). The task required to discriminate the identity of the lateralized target letters by pressing one of two lateralized response keys, in a corresponding or a non-corresponding position with respect to the target. Results showed that GCE was not modulated by the presence of the Mask, but it occurred in the No-Mask condition, confirming previous studies. Crucially, the GCE interacted with Simon effect in the Mask and Control conditions, though in different ways. While in the Mask condition the GCE emerged only when target and response positions corresponded (i.e., Simon-corresponding trials), in the Control condition it emerged only when they did not correspond (i.e., Simon-non-corresponding trials). These results indicate that people with face masks induce us to jointly orient our visual attention in the direction of the seen gaze (GCE) in those conditions resembling (or associated with) a general approaching behavior (Simon-corresponding trials). This is likely promoted by the fact that we tend to perceive wearing the mask as a personal safety measure and, thus, someone wearing the face mask is perceived as a trustworthy person. In contrast, people with a patch on their face can be perceived as more threatening, therefore inducing a GCE in those conditions associated with a general avoidance behavior (Simon-non-corresponding trials).

Face Mask Reduces the Effect of Proposer’s (Un)Trustworthiness on Intertemporal and Risky Choices

Previous literature suggested that individuals increase temporal and risk discounting at the presence of a proposer whose face is perceived as untrustworthy, suggesting the activation of protective choice patterns. By the way, the COVID-19 pandemic has substantially transformed the way we interact with other people, even bringing us into situations where the face of the person making a proposal is not fully visible, because of the mask. With the current study, we aimed at verifying if the effect of proposer’s facial (un)trustworthiness on discounting behavior is modulated by mask wearing. In two different experiments, participants performed traditional delay and probability discounting tasks with masked proposers manipulated across trustworthiness levels. Results highlighted that, even after checking for subject-specific emotion recognition ability with masked faces, the presence of a masked untrustworthy proposer increases both delay and probability discounting parameters, although the effect is not statistically significant and smaller than the one detected at the presence of an untrustworthy proposer without a mask. These results suggest that the ability to perceive the proposer’s (un)trustworthiness is affected by the mask, with a consequent less strong effect of proposer’s (un)trustworthiness on choice behavior on both intertemporal and risky choices. Limits and possible implications are outlined and discussed.

Early Word Segmentation Behind the Mask

Infants have been shown to rely both on auditory and visual cues when processing speech. We investigated the impact of COVID-related changes, in particular of face masks, in early word segmentation abilities. Following up on our previous study demonstrating that, by 4 months, infants already segmented targets presented auditorily at utterance-edge position, and, using the same visual familiarization paradigm, 7-9-month-old infants performed an auditory and an audiovisual word segmentation experiment in two conditions: without and with an FFP2 face mask. Analysis of acoustic and visual cues showed changes in face-masked speech affecting the amount, weight, and location of cues. Utterance-edge position displayed more salient cues than utterance-medial position, but the cues were attenuated in face-masked speech. Results revealed no evidence for segmentation, not even at edge position, regardless of mask condition and auditory or visual speech presentation. However, in the audiovisual experiment, infants attended more to the screen during the test trials when familiarized with without mask speech. Also, the infants attended more to the mouth and less to the eyes in without mask than with mask. In addition, evidence for an advantage of the utterance-edge position in emerging segmentation abilities was found. Thus, audiovisual information provided some support to developing word segmentation. We compared 7-9-monthers segmentation ability observed in the Butler and Frota pre-COVID study with the current auditory without mask data. Mean looking time for edge was significantly higher than unfamiliar in the pre-COVID study only. Measures of cognitive and language development obtained with the CSBS scales showed that the infants of the current study scored significantly lower than the same-age infants from the CSBS (pre-COVID) normative data. Our results suggest an overall effect of the pandemic on early segmentation abilities and language development, calling for longitudinal studies to determine how development proceeds.