Content
summary Summary

Much of social media is about beauty. A new Beauty AR filter from TikTok adds to that.

Beauty AR filters have been a controversial topic for years, especially on Instagram and TikTok. They conjure up perfect makeup, emphasize cheekbones, turn a hooked nose into a snub nose. Everything that the ideal of beauty demands. This is made possible by the increased processing power of smartphones and huge advances in computer vision.

"This is not healthy"

TikTok's latest beauty filter has reached a new level of realism: it captures the entire face and is so robust that it remains flawless even when you pluck your artificial eyebrows or wipe your hand across your face.

Many TikTok users are amazed by the before-and-after effect. One user, as she deactivates the filter and her real face reappears, says that her insecurity is about to skyrocket, and she comments that "this is not healthy".

Ad
Ad

According to TikTok user zoe_george_, the filter is "perfect" - but also dangerous. Many girls may not realize that others are using the filter and may follow the standard of perfection it sets because they think that's what people look like.

User "memo_akten" describes in detail the changes the filter has made to his face: It widens the forehead, lowers the eyebrows, slightly raises the outer corners of the eyes, makes the nose narrower, the lips fuller, the upper lip shorter, and the chin stronger.

In general, the filter affects various faces differently. In the case of "memo_akten", the effect is quite subtle.

Criticism of beauty filters

In particular, beauty filters that manipulate facial features have been criticized for distorting reality. They can have a negative impact on the self-perception of both users and viewers.

A study published in June 2019 in the Journal of the American Medical Association showed that such beauty face filters can increase interest in cosmetic surgery. As a result, Instagram banned filters associated with cosmetic surgery in October 2019.

Recommendation

In December 2021, a marketing study on the effect of makeup AR filters, i.e. without physical facial manipulation, examined self-perception before and after using a makeup AR filter compared to a traditional mirror image.

The result: The initial level of self-esteem strongly influences the effect of AR filters. Subjects with high self-esteem experienced a 44 percent greater gap to the ideal state after using the filter than after viewing their mirror image. In contrast, subjects with low self-esteem reported a 16 percent smaller gap to the ideal state after using the AR filter.

How does this fit together?

According to the researchers, subjects with low self-esteem felt more "beautiful" with the AR filter, i.e., closer to an ideal. The filter showed them an attractive, possibly attainable alternative state of their own selves.

Ad
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.
Ad
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.

High self-esteem subjects, on the other hand, were more likely to feel unsettled by the visually realistic intervention in their natural appearance.

In a follow-up survey, the researchers found that the AR filter had increased the discrepancy between the participants' desired and actual appearance, ultimately leading to a reduction in their self-compassion and tolerance for their own physical imperfections.

Support our independent, free-access reporting. Any contribution helps and secures our future. Support now:
Bank transfer
Summary
  • More realistic augmented reality filters are possible thanks to huge advances in computer vision.
  • So-called beauty filters model and make up the user's face according to common beauty standards.
  • TikTok users are amazed by the photorealistic quality of a new filter - and warn against its use.
Online journalist Matthias is the co-founder and publisher of THE DECODER. He believes that artificial intelligence will fundamentally change the relationship between humans and computers.
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.