Mirror selfies are one of the artefacts of postmodern Western society, which may appear as a flat and banal research element, but its prevalence, popularity, way of grasping speak of the contemporary art of popular culture on social networks. Selfies are a global phenomenon that we perceive as a sign used for self-presentation, communication and self-identification. Although one of the values of postmodernism is narcissism, personalization and individualism, under the influence of society and the economy we have paradoxically reached the mainstream, where the uniqueness of the personality is lost, which is represented not only by the physical body and expression, but also by material objects. Clothes and home are the strongest XXX of our identity. But in what ways are individual subjects and their interactions projected on social networks? In what ways does visual content shared on social networks influence the shaping of residential interiors? Due to the tremendous pressure from society through digital media, the way we furnish our homes and dress is more of a form of simulation that can lead to an unsatisfactory way of life. As we let masses of people into our wardrobes and living interiors via mobile devices, the line between private and public space is blurred. The following research deals with the relationship between fashion, interior and identity in the form of mirror selfies, where images from the Pinterest platform were visually analyzed using the big data analysis method.
IDENTITY, POSSESSIONS and SIMULATION
Our fragile selves need to be supported by something material and stable that helps us identify ourselves. "...Our possessions play a significant part in our identity and reflect it." If we buy an object in a store, we do not only have control over it, but above all it has control over us, because it has convinced us to buy it with its properties. This is precisely why our home interiors and clothes are the strongest media for materializing our identity, as we are in contact with them every day and they are with us. The role of the objects we own is not only to satisfy our physical and psychological needs, but they are also a material medium for preserving memories and feelings from the past and carry significant emotional value. Such value is different for each person based on the subjective attachment to the object, the associations associated with it. If we accept the proposition that the things we own are extensions of our own selves, then our home and what we wear are our second skin. The twist then comes when we are surrounded by objects that do not resonate with our identity and our daily needs.
Based on the theory of the narrative self-constitutional view (a term proposed by Marya Schechtman), which we understand as the continuous telling of a story with regular updates. A practical example is the regular sharing of content on a personal profile on social networks. Naturally, we buy things with which we would tell about ourselves and, ideally, get positive feedback from those around us. This need for self-presentation through non-material forms such as opinions or experiences and material forms such as daily necessities, food, clothing or household equipment, then logically affects our way of consumption and the way we then show it to those around us. Not only the objects in question, but also their location, the way they are photographed and the context in which the objects are placed allow users of social networks to see reality only as the author thinks. This leads to constructions of hyperreality, in which we cannot assess their authenticity with certainty, and therefore we perceive them as reality, which is, however, simulated using empty characters. This theory of simulation and simulacra is introduced by media theorist Jean Baudillard.
MIRROR SELFIES
Although mirror selfies may seem like a new genre of popular art, their tradition goes back a long way in history, where the depiction of oneself in a mirror or mirror surfaces can be recorded in the history of art in works such as Arnolfini Couple by Jan van Eyck, Las Meninas by Velazquez, Parmigiano Self Portrait , Self Portrait by Johannes Gump, etc. However, the current form of these self-portraits is different and quite specific. Although mirror selfies are a space for experimentation, the vast majority of images are tied to repetitive compositions, color tuning, format and concept. Reality can then be distorted, of course, by certain filters, cropping or other forms of editing. Although taking selfies is a social means of communication and self-presentation, women are the gender that takes pictures more often. At the same time, women are considered to be the main consumers in the field of fashion and interior furnishings. That is why they became the research sample in this study.
VISUAL ANALYSIS OF MIRROR SELFIES
The key analysis of the visual aspect of mirror selfies was provided by the Faculty of Applied Sciences at the University of West Bohemia in Pilsen, which is offering interdisciplinary cooperation. Under the guidance of the head of the Department of Cybernetics, Ing. Marek Hrúz, PhD. together with PhD student Ing. They started consulting Jiří Vyskočil on the possibilities of artificial intelligence.
My intention was to confirm or refute the hypotheses that I had prepared from earlier observation. It was Big Data Analysis through artificial intelligence that was supposed to provide me with accurate data.
Before the analysis itself, I set the questions that interest me and that are relevant for the research:
- Does the color match the clothes and the interior when taking a selfie?
- In what rooms are the photos taken?
- Does the composition repeat itself in the photos? (By composition in this context I mean the placement of the person, the placement of objects, the line of the mirror)
- Do the poses repeat?
- What percentage of the characters have their faces covered?
The input for the analysis were images downloaded from the Pinterest platform, which I obtained using the method of random sampling algorithm. Over the course of two weeks, I used the Pinterest website, where I entered the term mirror selfie woman and randomly downloaded images, both on the default page and through other clicks. In this way, for scientific purposes, I downloaded the mentioned 1572 photos of women taking pictures of themselves with a mobile phone in a mirror.
PhD candidate and colleague Ing. Jiří Vyskočil performed individual methods, which are described in the appendix, where the output was the same image with marked joints on the figure and detailed data below it. I obtained all the analyzed data through Excel tables, which I further processed in written and graphic form.
Each image that first went through Image Segmentation, Face Overlap, and Pose Estimation processes yielded the following data:
Image resolution defined in px.
Position of the figure from the edges defined as coordinates, in px.
The area of the person in the image is defined as a figure in %.
The position of the figure, which was assessed as standing or sitting with the help of the joints.
The area of the mobile phone in the image is defined as a figure in %.
Occurring objects in the image and their number.
The common attribute of the photographs was the position of the figure in the picture. The visualization below captures 9 segments of the image (created after division using the rule of thirds) with the percentage representation of the character in a specific segment. In most cases, there was a figure, and in the middle sector of the frame it usually occupies 70%. The corners of the picture are minimally occupied by the figure. The saturation of the color segments corresponds to the frequency of occurrence of the position. The most common position of the figure is therefore the center itself and the spaces above and below it. For greater clarity, I am also attaching a photo where the grid is transferred to the image itself.
As mentioned above in the methodological section, if the face is covered by the mobile phone at least 50%, we consider it covered. It follows that 965 female figures have their faces uncovered, while 607 do. 61.4% of women have their faces covered by a mobile phone. By comparing the position of the joints, it could be defined whether the figure is sitting or standing. Pictures when the selfie is taken while standing are significantly more frequent, occupying a full 84.1% of all analyzed. 40 photos were not recognized due to an error. The popularity of standing poses was verified both by means of a questionnaire and by means of this method.
First, I derived the following using mathematical formulas to understand in what scale specific objects occur in images:
The bed occurs approximately every 2nd frame.
A chair or sofa can be found in almost every picture.
The table occurs approximately every 3rd frame.
A bag, suitcase or other luggage appears on every second picture.
A vase occurs every 6th frame.
Plants are found in almost every picture.
Doors occur almost every 3rd frame.
Stairs occur every 37th frame.
Closet occurs in almost every 4th frame.
Unfortunately, we cannot consider these claims to be 100% accurate, as some objects, usually flowers or chairs, appear several times in the images. Even so, we can define the most frequently occurring furniture and interior accessories in the analyzed mirror selfies. They are plants, seating furniture and beds. It is the appearance of the bed as the third most numerous piece of furniture that confirms the claim about publishing privacy in social media. As part of the questionnaire, which I mention in the subsection Mirror as a means for mirror selfies, 30% of respondents said that they most often create mirror selfies in the hallway or hall, as well as in the bedroom and bathroom.
One of the other aspects analyzed is the location of objects. Primary data from the sources of Ing. Jiří Vyskočil converted it into a graph that captures the position of 9 objects in relation to the character depending on the frequency of occurrence. To define the sectors, we used the division of the image into thirds, which is a basic rule used especially in photography and painting.
This part of the analysis aimed to generate two color palettes for each image, which will consist of the five most common shades for the interior and the figure separately. Skin color is part of the character's color scheme, which is why the colors of the clothes correspond to only 4 shades in reality.
From the initial analysis I made, it is clear that the color shades are often repeated and the range of color palettes is usually not wide. An analog color scheme uses combinations of stronger or weaker tones of related colors that are adjacent colors in the Oswald circle spectrum. This applies to both the interior and the clothing, however the color schemes of the room are overwhelmingly analogous, occasionally supplemented with a contrasting color to a small degree. The color scheme is dominated by shades of brown (from beige to dark brown). There are also muted and pastel shades of blue, pink, purple and green. The proportion of occurrence of complementary colors and accents is very small, almost negligible. A wider spectrum of colors is more likely to be found in clothing.
The main, most widespread color for the outfit occupied an average of 36% of the entire figure, for the interior it was 42%. Usually these two colors were in neutral tones of brown, gray, beige or dark blue. In 1,005 cases, the main colors of the interior and clothes were adjacent colors or even in the same tone, but darker or lighter than each other. The more prominent color shades (accents) were usually the last two color shades, occupying an average of 12% and 7% in clothing and 9% and 3% in interiors.
I decided to verify the recurring patterns that I observed in images on the Pinterest platform and that were confirmed by data and analysis through artificial intelligence methods in a third way. I used ChatGPT to generate 150 mirror selfies to find out whether the images would correspond to the most common compositions and color schemes and whether individual interior elements and poses of the authors would be repeated. The result was surprising, as the vast number of images turned out exactly as if they had been downloaded from Pinterest. Although these photos have not been analyzed again in the same way, from mere observation I note similar color schemes, the placement of furniture elements, repetitive poses and the moment of covering the face with a mobile phone. The style is mostly minimalistic with bohemian elements, light colors and natural materials that give the room a cosy, casual and pleasant feel.
The research confirmed my previous hypotheses about the mainstream, which manifests itself in identical compositions, repeated furniture elements, uniform color schemes and the type of rooms, which were most often bedrooms. Taking and sharing mirror selfies on social platforms contributes to the harmonization of clothing and interior. The loss of the authors' uniqueness was also confirmed by the poses that tried to appear casual, but they are artificially constructed poses for the purpose of taking a photo. An interesting moment that gives room for further research in the field of photography and psychology is the frequent covering of the face and especially the eyes with a mobile phone.
From the photographs and from the data of my research in the form of a questionnaire, it is evident that stylization, or changes in the visual appearance of the room, often occurs before the photograph is taken. This fact is a stimulus for discussion, whether the focus of attention on a specific space is changing? Bedrooms used to be a purely private part of the home, which was not accessible to everyone, unlike social rooms such as living rooms or dining rooms. Now, however, the bedroom is exposed to the entire online world. Can it affect the focus of designers, who most often deal with the design of living rooms, kitchens and dining rooms? Will we want to invest more energy and finances in the bedroom for self-presentation, or in common areas? Or, on the contrary, the goal will be to buy cheap but fashionable bedroom equipment for the possibility of seasonal change.
As I already indicated, the perception of privacy is also changing in interior design due to the influence of social networks. In addition to images of bedrooms (and bathrooms), the claim is supported by the popularity of bare clothes and shoe cabinets, a furniture element used primarily to display clothes in fashion stores. At the same time, department stores often incorporate elements from residential interiors into their branches to achieve comfort and a pleasant atmosphere. Is it possible that there would be further adoption of elements from these two spaces? Is it normal for us as humans to show the world our intimate space such as bedroom? Or is it becoming normal?