You already know that your Instagram account says a lot about you. It may also be able to tell your mood, according to research published ahead of print through the Cornell University Library.
Using Instagram data from 166 people (a mix of healthy and depressed individuals), researchers from Harvard University and the University of Vermont showed that computer analysis of a person’s photo filters, colors, and the number of the people in the photo can be used to successfully screen for depression.
One portion of the study applied computer photographic analysis to over 43,000 photos and assessed colorization, vividness, brightness, and facial recognition. The other portion was conducted via Amazon’s Mechanical Turk, a crowdsourcing internet marketplace where workers do human intelligence tasks that computers are currently unable to do. On this platform, people were asked to rate photos on a scale of one to five based on how happy, sad, or likeable the person in the image seemed. The research found that the computer was able to predict which users were most likely to suffer from depression better than humans.
The number one filter linked to depression, you might ask?
Although the research showed that depressed users tended to avoid using filters in their photos altogether, when they did, they overwhelmingly favored the “Inkwell” filter. On the other hand, healthy users were found to heavily favor the “Valencia” filter. Interestingly, depressed users were also most likely to Instagram photos of individuals and recieve more comments, while healthy users tended to recieve more likes.
The research from this study, which has yet to be peer-reviewed, could be used in the future to build self-diagnostic programs on mobile apps and prompt users to seek treatment earlier than they otherwise might. Until then, you may want to consider making “Valencia” your go-to filter.