“Heyy!! You have been online the whole day on Instagram and posted countless stories but at the same time you are not picking up a single call”. She said in one breath. “What is your problem?? And stop posting such sad stories ok”. Often this is what we do. We just consider the posts are just something we share as stories and that is it. But you know what, it isn’t. It is a lot more than that.
Man is a social animal. Yes, that is what we are called. But this social animal is too picky about what it is social. Happiness yes definitely but when it comes to sadness he prefers to be alone. But this sadness reflects somewhere.

With AI entering into every field known and unknown how can it stay back here??
Researchers from the University of Vermont and Harvard University built a tool that can Detect Depression Using Artificial Intelligence. Well yes! Our social animal needs this now.

What is depression?

Depression is a feeling of severe loss of hope and dejection. In this state of mind a person loses interest in everything, all he does is stays by himself with no communication with anyone. Previously when people used to actually talk to each other the cases of depression were less compared to today when we say the social media has given us a platform to express.
Once a person showed signs of depression we didn’t really know what depression was. What we thought is the person is ill and a doctor was called. Now who has the time to sit and observe the change in a person’s behavior, we just claim it to be a tantrum and walk off. This is where the actual problem starts. ‘No one cares for me’, ‘I don’t mean anything to anyone’, ‘who cares what I feel’, ‘everyone is happy with his or her own lives why do I tell them my sad story’. Shocking?? Heard of these things lately?? And this my friends is depression.

AI to detect depression?

Sadly, yes. Today we need a software for everything so why not to detect depression?? We are all working to make machines with capabilities to be humans and this brings us a step closer. Our social media says a lot about us, including the knowledge of our mental health and if we need medical help. Scientists have prepared an algorithm that flags a few signs from our Instagram posts and detects depression.

From the studies and researches, we can say that people who tend to be depressed posted photos that are in the shade of blue, grey, in darker colors or black and white. unlike healthy people who post colorful pictures. Depressed pictures contain fewer or no people and also lack filters.

How does the algorithm work??

The algorithm considers a few details of the user. These include,
• How often the person is posting. (Depressed people tend to post less)
• The shade of the post. (Found to be bluer, darker or greyer)
• The number of likes the person gets. (Lack of social interaction leads to this)
• The type of posts the person likes. (posts depicting sadness seem to be a priority)
• The filter used. (generally, no filter or inkwell is used)
• The number of people in the post. (fewer people, again due to reduces social interaction)
• Sadness selfie. (a sad person posts a selfie in the dark with a sad caption)
And a lot more. Then these symptoms are flagged and a final decision is made regarding the depression of the person. Though the algorithm is used to Detect Depression Using Artificial Intelligence only 70% of the time it is a lot more than the diagnosis done by doctors. A study also shows that doctors detect depression only 42% times.


This algorithm or this tool definitely isn’t the ultimate solution to depression detection. But it can act as the breakpoint of the development. The further development can be done upon this baseline. This will definitely help us detect depression at its early stage and take action before it is too late to act.

Read More Tech News On Techie Scoops. Also, Follow Techie Scoops on Instagram for quick scoops.

Motion Capture – A Peek Into Tech Used In Animation

Previous article

Internet In Flight: Now Access It While Flying Over India

Next article

More in Tech News

You may also like


Leave a Reply