Keywords

1 Introduction

Selfies play a major role in self-presentation on social media sites. Visual social media platforms, such as Instagram and Snapchat, have grown in popularity over the past fifteen years especially among 15–25 year old members of generation Z. Since Instagram and Snapchat introduced augmented reality (AR) filters in 2015, these filters have become a popular widespread feature for taking selfies. 600 million people use them each month on Instagram or Facebook, 76% of Snapchat users apply them every day [9]. Some of these filters are silly, such as cat ears or fantasy characters, while others allow users to digitally alter their face to conform to specific beauty standards. These augmented reality beauty filters (ARB filters) do not only apply digital makeup, but they go beyond that by deforming the user’s jaw and nose, expanding their eyes and lips, and smoothing their skin [2]. Prior to the development of ARB filter, selfies could only be enhanced by retroactive photo editing. Now, facial enhancement is achieved via AR filters that adapt to facial features in real time. This takes the interaction with the beautified self to a different level. While taking a selfie, the user moves and their beautified self moves with them. There seems to be a gap in the literature when it comes to the specific effects of the process of taking a selfie with AR beauty filters in contrast to retroactively editing a static image. This study aims to shed light on these differences and the effects of AR-beautification on self-perception.

2 Background

2.1 Augmented Reality Beauty Filters

Augmented reality is a technology that combines reality and digital information by overlaying digital content on objects, humans or places in the real world. ARB filters are automated photo editing tools using artificial intelligence and computer vision to detect facial features and modify them [11]. Beauty filters are developed by individual creators as well as cosmetic brands and retailers. While some beauty filters are merely for applying make-up, the majority of beauty filters alter the contours and shapes of the face. Most beauty filters modify their users’ facial features in a similar way. It’s what the journalist Jia Tolentino points out as the Instagram Face characterized by poreless skin, high cheekbones with catlike eyes, a slim nose and full lips. It is a face that is “distinctly white but ambiguously ethnic” [14].

A study by Rosalind Gill, a professor of Social and Cultural Analysis, conducted in the United Kingdom in 2021, found that 90% of young women either apply beauty filters or edit their photos before posting them on social media. According to Gill, beauty filters and photo editing contribute to a society where young people constantly feel scrutinised and judged by their peers [7]. The way AR technology is used in social media raises a number of other ethical concerns, ranging from the promotion of plastic surgery to facial recognition to questions about the nature of AR. Behr et al. [2005] [3] formulated the following four risks of XR (virtual and augmented reality) technology: (i) motion sickness; (ii) information overload; (iii) intensification of experience and (iv) cognitive, emotional and behavioral disturbances after re-entering the real world after the XR experience. XR technology is highly persuasive, and that is where the risks of ARB filters lie. Since the “reality” of the face is experienced through the device, virtual aspects may become indistinguishable from the real. For vulnerable groups such as children and adolescents, discriminating between the real and the virtual can pose an even greater challenge. The embodied experience of virtually altering one’s face can create confusion in people about their real bodies and lead to a kind of body dysmorphia [13]. Finally, the immersive nature of AR technology could change perceptions of the self in a different way than normal photo manipulation.

2.2 Effects on Self-perception

Motivations. The motivations for self-presentation on visual social media platforms such as Instagram, Snapchat or TikTok are multi-faceted. They are associated not only with the drive to present the ideal self, but also with presenting the true self and transforming the self [9]. Those who use AR filters to depict an idealised version of the self often want to come across as more beautiful or cool. This fits with the user behaviour of many, where only content that shows the best moments or looks is uploaded [12]. For those wanting to present their true self, AR filters can be tools to experiment with different aspects of the self. Despite the digital modification, representation through AR filters can be congruent with the user’s true self. Finally, filters can be used to transform and discover new aspects of the self. This allows individuals to engage in cognitive processes (thinking, hoping or fearing) about who they could be [9]. When users engage with filters for entertainment and fun it is often associated with escapism. Users may feel like they are mentally transporting themselves to imaginary worlds. According to Javornik et al. [2022] [9] the impact of AR filters on mental health and well-being depends on the underlying motivations for their use. Those who use AR filters as a form of entertainment, social interaction, or as means to transform their digital selves often benefit from the technology, and a positive effect on the mood can be observed. However, using AR beauty filters to idealize the self and fake one’s image often leads to lower self-acceptance and has negative effects on users’ self-perceptions.

Self-discrepancy. When we look at ourselves through the lens of our phone camera or in the mirror, we may see the same image. Yet AR technology allows us to overlay a filter on the image on our phone in real time. This creates an obvious discrepancy with what we see when we look in the mirror, where no such filters can be applied. The negative effects resulting from this experience can be explained with the self-discrepancy theory (SDT) by Higgins [1987] [8]. According to Higgin’s theory there are three domains of the self: a) the actual self, a representation of the attributes a person thinks they possess; b) the ideal self, representing the attributes that someone would like to possess and c) the ought self which represents the attributes that someone believes they should possess. A perceived discrepancy between the actual and the ideal self can lead to feelings of disappointment and dissatisfaction as one feels that one cannot live up to one’s own beauty aspirations. Discrepancies between the actual and the ought self can also result in negative emotions and in the fear of negative outcomes. According to SDT an individual is motivated to align their actual self with the ideal self and the ought self. In their work on the effects of beauty filters on the self-image of Saudi women Alsaggaf [2021] [1] stated that the use of beauty filters on Snapchat leads to a contradiction the user feels when seeing the actual image compared to the ideal image using Snapchat beauty filters. The negative emotions following this perceived contradiction align with SDT.

Extended Mind Theory. One risk associated with AR is its intensity, which can make it difficult for users to distinguish between the real and the virtual world. Using Clark and Chalmer’s Extended Mind Theory the human and the AR application could be considered a coupled system. The distinction between the real and virtual world becomes difficult because the virtual extends the boundaries of the mind, rather than being external to it. In his book “Reality+” Chalmers [2022] [4] argues that AR affords us the opportunity to expand and augment our minds. AR glasses that project virtual items into the actual world to aid in navigation augment both the physical world and the mind by enhancing mental capacities [4]. Similarly ARB filters may extend our mind by allowing us to visualise what a “better” version of us might look like.

3 Method

3.1 Participants

8 people participated of whom 4 identified as female, 2 as non-binary and 2 as male. Their age varied between 24 and 33. All participants were White and of either German or U.S. origin. Convenience sampling was used for recruiting the participants.

3.2 Materials

Each participant used a computer that had Zoom installed. Additionally the participants used their smartphone and the Instagram app that was installed on it. Six participants interacted with a filter called “Mary Phillips” created by Sophie Katirai that adapts the face to female standards of beauty. Two participants interacted with a filter for male beauty standards called “Men’s Beauty” created by Jason Emer. These two filters were chosen because of their popularity. Both creators are famous influencers whose filters are frequently used.

3.3 Procedure

At the beginning of the Zoom interview participants were asked to hide their own self view on Zoom, so that their own reflection on Zoom wouldn’t distract them. They were then asked to have their Instagram app with the filter ready and also have their front camera open, so that they could switch between the two applications on their phone. There were two parts in this experiment. In the first part participants were asked to apply the filter and take a frontal picture of themselves without making any movements. They were then asked to look at this static beautified image. In the second part participants were asked to apply the filter and move around and look at themselves from all kinds of angles. After and before each part of interacting with the static or moving beautified self they were asked to switch back to their front camera. In the second part when they moved with the filter participants were asked to take a selfie with the filter. Later they were asked to take a selfie without any filter. Participants were asked to indicate when they started and finished with taking a selfie. That way a time stamp could later be identified in the transcript.

3.4 Interview Questions

A semi-structured interview was conducted with most questions belonging to one of the three categories: a) the emotional experience, b) self-perception and c) selfie behaviour. Questions were asked during and after the interaction with the ARB filter. Category a) includes, for example, questions like: “How do you feel about yourself when you look at yourself with the filter?”, “What kind of emotions did you experience in the moment when you switched from the filtered version of yourself to the no-filter version?” Questions belonging to category b) comprise “How much do you identify with this filtered version of yourself?”, “Which facial features are modified by the filter?”, “Is there anything about your face you notice in a different way now?” Category c) includes: “What kind of movements do you do when taking a selfie?”,“Were the movements you did when you were taking a selfie the same with and without the filter?” Furthermore participants were asked whether they experienced a difference between interacting with the static or moving beautified self.

4 Analysis

(See Fig. 1).

Fig. 1.
figure 1

Left with the ARB filter, right without a filter.

Full size image

4.1 Self-perception While Using the Filter

Feeling Disconnected. All participants described their experience of interacting with the filter, either static or moving as somehow fake. Four participants used the word “artificial” to describe their beautified self. Often the artificiality was the reason why participants did not like the filter or could not identify with the filtered version of themselves. One participant described her experience like that: “I definitely feel like it makes my face look so much more artificial. It makes me feel like I’ve gotten plastic surgery”. Most participants described a feeling of disconnection when looking at themselves with the filter. Two participants used the word “alienating”. One participant even experienced anxiety: “Seeing myself in that altered, and augmented way kind of gives me anxiety. It’s just weird because I never use these filters and now I see it as like: ‘I actually could even look better’ kind of. I don’t know if this is necessarily better. But just the possibility of it kind of triggers a weird feeling, a kind of anxiety maybe”. Five participants said that they feel like a different person with the filter. One participant said “I feel like I’m looking at a Disney movie version of myself”. Another participant said: “It just kind of reminds me of a generic Instagram model face. It’s just sort of a very trendy cat eye, blush, makeup, kind of person”. A third participant described her experience like that: “I think I less identify with myself and more with like Instagramers you know like influencers”. Two other participants attributed this different person that they saw even with a different kind of personality. One said “I don’t think anyone looks like that. Or if they really did I feel like I wouldn’t like them”. Another participant articulated “If I would look at a picture like this. I wouldn’t think it’s a sympathetic person. It’s very artificial”.

Positive/Negative Aspects of the Filter. All participants liked certain aspects of the filter. For some it was the makeup, for others it was the enhancement of the lips or the smoothing of the skin. Most participants liked that the dark circles and wrinkles under their eyes disappeared, making them look more fresh. Two female participants noted that the filter made them look more young, one said: “With the filter when I move I actually feel very cute. I think my lips are very big and the head is small. It looks really a bit like a child”. Nearly all participants reported that the filter might “objectively” make them more pretty. Overall however, all of them rejected the idea that the filter made their face look more beautiful than in real life. One participant said: “Its kind of like, maybe adapting my face to beauty standards or like a perception of how male beauty works that I’m not subscribing to”. One participant noted that she did not like the way the filter shapes the eye into a cat eye form: “I know that is actually from like a racist idea of beautifying Asian features, so I don’t like that”. Two participants complained that the filter erased their freckles, a feature they both like about their face. Most participants found the filter too strong and noted that they would like it more if it was more subtle.

Getting Used to the Filter. Most participants described that the longer they looked at themselves with the filter (either moving or static) the more real it felt since they got used to it. One participant said: “Like the first second I was like okay that’s really not me, but the longer I look at it I would say it looks kind of similar to me”. Another participant said: “The longer I stare at it the less absurd it looks. And the more I could see like: ‘okay yeah I can look like that’ or it doesn’t look that unnatural”.

4.2 Emotional Experience After Using the Filter

After Interacting with the Static Beautified Selfie. When going back to the unfiltered front camera three people described their emotional response as more positive than negative. Looking at themselves again without a filter made them realise that they do actually like their normal face more than the filtered version. One participant said: “My first reaction was that I’m surprised that I don’t look as bad as I thought I did. I don’t see my flaws or they’re not glaring out”. Three participants experienced negative emotions when they switched back to their front camera. One participant said: “I sense some sort of downgrade.”, another said: “Going from a beauty filter that has a lot of makeup to a no makeup face, that’s a little jarring”. Most participants pointed out that certain facial features became more visible to them after looking at the beautified selfie. One participant said: “I do see things that I didn’t see before like I realise more now what the differences are. Like my wrinkles around my eyes, and maybe my nose and my lips. I don’t feel like before, I do feel like it’s changing that. Because before I felt pretty positive about myself”.

After Interacting with the Filter While Moving. Seven out of the eight participants described that they were experiencing a negative emotional reaction in the moment when they switched back to their front camera. Two participants described their emotional response with a feeling in their heart. One said: “I do felt like some heaviness in my heart somehow. Yeah, it’s this kind of unsettling feeling. Even though I’m not sad, but there’s this kind of anxious feeling for sure.”, the other participant said: “I definitely kind of felt like a heartdrop. Like: ‘this is what I look like’ and then having to really quickly smile and remind myself that I look great regardless. The more I looked at the filter the more I got used to seeing it and switching back I had to very quickly go through those emotions”. Again most participants seemed to notice certain facial features more than before applying the filter. One participant described it like that: “I can definitely notice how different my jaw line is between the filter and no filter. And that doesn’t make me feel great. And so looking at my face now, I just noticed my jaw, a lot more. I look so much more round without the filter”. Three participants stated that they were disappointed when they saw their face without a filter again. They used expressions like “downgrade” and feeling “underwhelmed”, “disappointed” and “less enthusiastic” about their physical appearance. One participant said: “I feel like a weight on my body. I feel more weak now”.

4.3 Selfie-behaviour

One objective of this study was to explore whether people behaved differently when taking selfies with and without the ARB filter. The transcript disclosed the time it has taken participants to take a selfie. The time it took them to take a selfie with the filter ranged between 3 and 10 s, whereas the time it took to take a selfie without the filter ranged between 5 and 20 s. Six of the eight participants took less time with the filter than without. Half of the participants indicated that taking a selfie with an ARB filter is easier than taking a normal selfie, because the filter allows them to take the selfie from a greater variety of different angles. One participant said: “I think the filter is a little bit more forgiving, in the sense that there are more angles where my face looks good. So, I can use maybe an angle that I normally wouldn’t go for on a selfie”. Three participants said that it didn’t make a difference for them in their selfie behaviour whether they applied a filter or not. One person said that she felt uncomfortable being watched while taking a selfie: “I feel like my selfie taking behavior is really influenced by the fact that I’m being watched. I was really like: ‘okay I really want to get out of the situation’. Otherwise maybe I would have taken more time”. One male participant reported that the filter did not always settle properly on his face. He said, that this is why it took him longer to take a selfie with the filter, since he needed to first find an angle where the filter is actually working.

4.4 Difference Between Static and Moving

Technology Failure. Most ARB filters on Instagram adapt the user’s face to female standards of beauty. In comparison there are only very few filters for male beauty. The filter “Men’s Beauty” used in this study worked a lot less well then the female beauty filter. One male participants who used the filter reported that the filter does not adapt quickly enough to his movements, resulting in a strange feeling and showing him that the filter is not real. He described his experience the following way: “I think it’s just like, funny somehow. And it’s ridiculous too. So, as I see this filter applying and then not applying or not being like set the way it has to be on my face I identify less with it. It’s like when I’m moving and then my eyebrows are kind of in my eyes. So that makes me realise it’s a filter and it’s not real”. This participant reported that looking at the static beautified image had a stronger effect on his self-perception than moving with the filter, due to this technology failure. The other male participant decreased the intensity of the filter during the experiment which improved the functionality and prevented the filter from not properly settling on his face. This aspect of technology failure did not occur to the same extend with the female and non-binary participants who used the female beauty filter. However, one female participant had a similar feeling of confusion when one time the filter did not set properly: “Just one thing, which is funny because I looked at myself and then I took a sip of coffee, and then it didn’t work anymore. So I was like really confused for one second, I was like: ‘oh I look nice’ and then suddenly my eyes looked like they normally do”.

Movements. Most participants indicated that the interaction with both digital beautifications (static and moving) affected them in some way. One participant said: “I think both are kind of jarring, and I think both change your perception of how you look”. Six of the eight participants found the experience of moving with the filter more intense than looking at the static beautified selfie. They noted that the beautification felt more real when they saw the filter adapt to their movements. One participant said: “When I move, the filter feels more natural. Maybe because movements are part of your identity and you can recognise yourself better”. Another participant regarded the filter more as a part of their face when they were moving: “When I was moving, I had more of a chance to think about the filter more. Before when it was static it was just like: ‘okay this is my face and this is the filter’ and it didn’t feel very meaningful that I was using the filter. It didn’t feel like it impacted me but then when I was moving around it felt more like it was a part of my face”. For another participant it was the possibility to see herself from different perspectives that made a difference: “I feel like being able to move from side to side made it look more like myself. I think it was especially visible from the front view that it doesn’t look like myself. The moving aspect made it look a bit more like me and I could identify more with it. Whereas with the frontal picture, I only had this one perspective and I was like okay this looks very different. But when I can move I do identify more with it for example when I’m seeing my nose from the side it looks more like my real nose”. Another participant also described the experience of moving with the filter as more intense yet she felt less like herself: “I think when I just used the filter while moving, I was kind of pretending to be someone”.

Time Spent Interacting with the Filter. All participants spent more time looking at themselves while moving then looking at the static beautified selfie. This time difference was not intended in the design of the experiment but more a result of how the participants answered the questions. All participants answered in more detail when they were moving around, it seemed like as they moved, there were more things they noticed about the filter. Moreover, they sometimes took longer to answer as they were moving and talking at the same time and sometimes seemed distracted by that. Participants spend on average 3.3 min looking at the beautified selfie and 6.2 min moving with the filter. Two participants noted that this time difference changed the impact the filter had on them. One of them said: “I think the only difference was just the length of the two things. I looked longer at myself while moving. The longer that I was interacting with the filter, the more I kind of got used to it and could accept it as myself. And I think that probably had a negative effect on me changing back to what I actually look like”. This time difference in the experiment must be considered as a potential factor that influenced the different experience people had when interacting with a static selfie vs. moving with the filter. However, one could argue that in real life people normally spent more time interacting with ARB filters on social media (since they spent some time moving with the filter before taking a selfie) than the time they spent looking at an already enhanced static selfie.

5 Discussion

5.1 Extended Mind and Self-perception

When switching back to their unfiltered front camera, many participants had negative emotions about their physical appearance. The interaction with the beauty filter made certain features of their face become more visible to them. Some participants were more aware of wrinkles, skin irritations, or the roundness of their faces - features that are usually not associated with beauty ideals. Their self-perception seemed to be affected in the sense that particular facial features stood out more to them after viewing their enhanced selves.

Six participants said that moving while using the ARB filter was a more intense experience than interacting with a static selfie. This aligns with theoretical considerations of extended mind theory. Applying Clark’s and Chalmer’s theory, the human and the ARB filter could be considered as a coupled system that extends the mind’s boundaries. This may result in an inability to distinguish between the actual and beautified enhanced self. ARB filters could serve as visualisation aids for the mental image of the ideal or ought self. Herein lies the distinction: the self and a manipulated static image of the self would not be considered a coupled system as there is no interaction between an image and the self. The cognitive process of imagining how one would look if they resembled the manipulated selfie would still need to be carried out by the individual. The augmented self however, seems real because it is not static, but reflects people’s movements. The AR beauty application and the self interact, facilitate a cognitive process and therefore become a coupled system. These considerations and the results of this study led me to hypothesise that AR facilitates a digital beautification process that could potentially have a greater impact on people’s self-perception than retroactive photo editing.

5.2 Self-discrepancy

The negative emotions the participants experienced when they switched back to their normal front camera could be a result of an increased perceived discrepancy between the ought self and the actual self. All participants disliked most aspects of the filter although it might adapt their face to certain beauty ideals. They rejected the idea that this beautified self improves their natural beauty. However, most participants did experience negative emotions after seeing their enhanced self. I assert that internalised beauty ideals of how one should look like influenced the participants more than their own ideal self. Higgin’s self-discrepancy theory discriminates between ideal-actual and ought-actual self-discripancies. According to the participants statements I argue that participants’ experiences can at best be described in terms of perceived ought-actual self-discrepancies. Five individuals reported seeing a different person when interacting with the filter. Surprisingly, two people assigned this different person unsympathetic character traits. According to Javornik’s studies on why individuals use filters, the participants may perceive the filter as a mask or costume that mentally transfers them to another world. Attributing their augmented self with different character traits could indicate that these participants identified less with the filtered version of themselves.

5.3 Sense-Making

Sense-making that happens through movements and interactions with the environment is often described in enactive and embodied cognitive science. I particularly found the statement of one participant interesting in that context: “When I move, the filter feels more natural. Maybe because movements are part of your identity and you can recognise yourself better”. The ability to move and interact with the filter seemed to assist her in recognising herself in the augmented version of herself. An analogy to the concept of dynamic touch first coined in ecological psychology by Gibson [1992] may be made here. Instead of mentally calculating the weight of an object by looking at it, dynamic touch describes an exploration with the hands. The work of the muscles that causes pressure and deformation to muscles and tendons offers information about the weight of the object that becomes available at the wrists [5, p. 105–135]. In enactivist terms the concept of dynamic touch aligns with sense-making activities [15]. Similarly vision is described by Alva Noë as a process that depends on interactions between the perceiver and the environment and involves movements [10]. Following this argument I claim that sense-making of the digital self can only happen when one moves and interacts with the ARB filter. This gives support to my hypothesis that AR as a technology could have a greater impact on self-perception than photo editing.

6 Conclusion

Overall this research adds to existing literature exploring the impact of beauty filters on users’ self-perception and body image. It calls into question current digital beautifying procedures by focusing on augmented reality that adapts to movements in real-time. The purpose of this study was to investigate how ARB filters influence people’s self-perception. Participants were first instructed to engage with a static enhanced selfie during the experiment. They were asked to look at themselves and move with the ARB filter in the second part. I was curious to see if the two components of this experiment had distinct effects on people. The majority of participants stated that moving with the filter affected them more than looking at the static enhanced selfie. It is not clear however if this effect could be observed due to the fact that people spend more time moving with the filter than looking at the enhanced selfie. Future studies could investigate the hypothesis that ARB filters have a greater impact on users than retroactive photo enhancement. Questions about how movements manifest feelings of identity could be additionally explored.