Instagram posts have been used to look for signs of mental illness amongst users. User settings, time of the day the photograph was taken, filters used and the number of people present in the frame are all pointers that have been to detect the problem.
In a study conducted by the SpringerOpen, named Instagram photos reveal predictive markers of depression, Instagram data of 166 individuals was taken, which were used to create tools to detect depression. This data was then used on Instagram posts of 43,950 participants.
Tools generated, used face detection algorithm to analyse Instagram posts. These tools detected if there were people other than the subject in the post, the number of likes and comments each post received along with the posting frequency. These figures solved the purpose of user engagement.
It was found that depressed people would comment on posts more; meanwhile, healthier individuals would like more posts. Also when it came to face count, depressed people were not only posting more pictures, the average number of people in each post was lower than those of healthier individuals, giving an idea of their social life.
Meanwhile, those with a high hue, saturation and value (HSV) in their photographs were less prone to depression as compared to the ones who had a lower HSV. The ones suffering from depression had pictures which had more blues, greys, and were darker.
Based on a hypothesis choice of colours and filters were also used to depict depression. While healthier individuals tend to chose Valencia filter the most, which tends to give a lighter tone to the picture, the depressed ones went for Inkwell to Crema which are known to add a black and white filter to your images. In fact, those who were feeling depressed, added less filters to their pictures.
While the study is able to detect some level of depression, one needs to realise that the sample size of the survey was about 166 people. So it cannot really be extrapolated for the entire user-base of Instagram. Moreover, the study did not take into account factors such as gender, race, economic situation or workplace situation.