
There appears to be a connection between the use of color in an Instagram photo and the mental state of a person.
A new algorithm seems to be 70 percent accurate when determining depression in people who use certain filters in Instagram.
Social media gives us a good insight into people’s life, maybe more than we sometimes want, but now researchers have discovered a machine that checks Instagram photos and can diagnose depression and other mental illnesses.
The new study conducted by Andrew Reece of Harvard University and Chris Danforth of the University of Vermont reveals the fact that there is a strong connection between the usage of color and the mental state of a person.
Any Instagram user will probably know that the photo-feed app provides around 23 different filters that you can apply on top of your photo. Researchers have concluded that the more a person uses one of the filters, the grayscale option called Inkwell, the more depressed they are likely to be.
The study was conducted on around 170 people who were asked to share photos from their Instagram feed and answer several questions from a standard clinical depression survey. After sharing some of their photos they were then asked to rate 100 photos from each person based on how sad, happy or interesting the photos appeared. Moreover, they had to comment on the filters chosen by others and the number of faces in each picture.
The researchers used the photos in a machine-learning algorithm that noticed the correlations between image properties and depression. This algorithm found that decreased saturation and brightness and increased hue predicted depression.
Also, specialists reached the conclusion that people with depression usually have fewer likes on their photos and tend to comment a lot on other people feed.
The findings showed that individuals who used less bright filters or no filters at all were more likely to suffer from depression and the accuracy of this prediction was more than 70 percent. Because of the high prediction, researchers believe that their technique can be used to discover early signs of mental illnesses.
What is certain for now is that an individual’s mental state is transmitted to social media and by using computational methods the elements expressed through photos can be easily identified.
Image source: Flickr












