Find out how boring your Instagram selfies are

2024 Author: Malcolm Clapton | [email protected]. Last modified: 2023-12-17 03:44

The general craving for photographing everything and everyone has led to the fact that the popularity of Instagram is constantly growing, and many of us are seriously concerned with the question of how to get more likes. Scientists from the Massachusetts Institute of Technology (well, who, if not them) even developed a special algorithm to help answer this question.

It's not just selfies, of course - it's about all the visuals. As a result of research by a group of scientists from the Massachusetts Institute of Technology, a special algorithm has been developed that allows you to determine how memorable your photo will be. That is, it helps to predict how likely people will remember or forget your photo. If you are curious how unique your own photographs are, then you can easily check it out here.

How the algorithm works

A few seconds after the photo is uploaded to the site, you will see something like a heatmap appear on top of it. It shows the areas that the algorithm considers the most and least interesting. Warm colors indicate objects that are more worthy of attention, cold colors - those that are of less interest. According to research, this algorithm perceives images as close as possible to how a living person would rate them.

Also, a scale and numbers are placed below the image. For example, if a photo has a memorability index of 0.91, this means that 91% of people who view it will remember the photo for another 100 seconds after viewing it. If you think this is not very long, just think about how fast we usually view our feed on social networks.

Practical utility

The algorithm makes it possible to improve understanding of how people perceive and process visual information, as well as learn how this information affects memory. The algorithm is useful because it allows you to understand which information is more important to people and which information they are most likely to forget instantly. This is similar to using the focus group method, which tells you what kind of reaction a particular image will cause.

During the research that was carried out at the stage of developing the algorithm, scientists made one interesting assumption that sheds light on the nature of human memory. They wondered what it would be like to show people only images with a high memorability index. Will any of them be forgotten anyway? It turned out that no, and this only means that if you continue to work on this issue, then in the future there will be an opportunity to significantly improve human memory.

If you think about the serious reasons for the need for such an algorithm, then you should not ignore its usefulness for marketers, filmmakers, as well as people who are somehow connected with the learning process. The algorithm will help to enhance the impact of visual materials and improve the quality of teaching resources. If we talk about medicine, then the developers of the algorithm think about how it can help health: for example, diagnose memory lapses or identify specific diseases using a heat map.

Plans for the future

The University of Massachusetts team plans to create a special application in the near future that will make the use of the algorithm more convenient for users. They will be able to improve the quality of their pictures and increase their impact on other people using a specially designed filter. It will not be intended to retouch something on the photo, but to make the photo more memorable. Also, the filter will help crop those places that are of the least interest.

In addition, the developers had the idea to complicate the algorithm so much that it itself removes uninteresting areas from the image. But it is too early to talk about this, since no one can guarantee that the output image will not look like a picture poorly processed in Photoshop.

So far, the algorithm has certain drawbacks, since it is not able to recognize all visual materials. It does not work well with the types of images that it has not encountered before. For example, now he works better with "live" objects and poorly knows how to recognize logos and emblems. This is exactly what the researchers want to focus on in the near future in order to improve the algorithm. As mentioned above, this will be of immense benefit to commercial development.

Last and most interesting

In the course of research, it was found that faces and body parts are more memorable to people than abstract pictures or landscapes. Beaches, sunsets and concept photography tend to fade from memory pretty quickly.