Photos you publish on social networks may determine if you have depression

Depression, as a common mental illness, is not easy to be clearly judged because of its invisible external features. Recently, however, researchers have developed a set of algorithms based on the difference in color preference among people with depression, and used “information” in photos and photos published by people to identify patients with depression.

Medical research shows that people’s favorite colors are influenced by their mood. When they are happy, they tend to favor bright colors. When they are depressed, they choose dark colors. According to this theory, two scholars from Harvard University and Vermont University developed a "depression recognition algorithm." In order to verify the accuracy of this algorithm, they chose to test the photos posted by users on the social network instagram. Researchers say that the accuracy of this algorithm for depressive patients is currently 70%.

Two scholars found 500 members of the Amazon Turk team with Instagram accounts to test. In order to confirm in advance which testers are likely to be depressed patients, they provided a questionnaire for clinical depression for all testers. In the end, 70 of them were identified as having clinical depression.

During the collection, researchers selected 40,000 photos from the instagram of all testers. In order to ensure the accuracy of the test, the photos of the health testers were selected from the latest 100 photographs. For the depressed patients, 100 photos were published before the treatment.

By using color saturation, contrast, and facial recognition software to test and analyze these photos, the researchers found that people with depression tend to publish photos in dark colors such as blue or gray, and there are also filter selections. Certain preferences, such as "inkwell" black and white filters, are more popular with people with depression. In contrast, healthy people prefer to use the high-brightness "Valencia" filter.

Recommended reading:

The Long March from Perception to Cognition: What Can a Computer "Read"?


New face recognition height: no face recognition!


The Application of Deep Learning in Face Recognition——The Evolution of Grandma Grandmother Model

Shear Beam Load Cells

Shear Beam Load Cells,Shear Beam Load Cell,Shear Beam Type Load Cell,Beam Loadcell

Xiaogan Yueneng Electronic Technology Co., Ltd. , https://www.xyeloadcell.com

This entry was posted in on