Harvard research lets Instagram detect signs of depression

It’s easy to fake a smile and say you’re fine so Harvard researchers look at how an Instagram post is composed to find signs of depression.  Image Inquirer.net

“A picture is worth a thousand words,” the old adage goes. Now it may also bare signs of clinical depression for Instagram users.

A research paper by Harvard researchers sampled 43,950 photos from 166 Instagram users using machine learning tools. They made use of algorithmic face detection, metadata components and color analysis to identify the signs of depression, reports Engadget.

Their study found that individuals who were more likely to suffer from depression had bluer, darker and grayer photos posted on their Instagram account. On the other hand, ratings that relied on photo attributes such as sadness, joy, etc., were weaker in predicting signs of depression.

In a sense, people who are “feeling blue” would tend to gravitate more to filters with shades of blue while those with a more upbeat disposition preferred warmer filters like Rise, Valencia and Earlybird. The study can be found here for more details.  Alfred Bayle