Need another opinion your latest selfie? A new bot trained to scientifically evaluate selfies is dishing out grades on Twitter.
Selfie Bot (@deepselfie) was created by Andrew Karpathy, a PhD student at Stanford University’s Computer Vision Lab, and is powered by a Convolutional Neural Network computer, a visual recognition system that is often used to identify people and things in images.
Getting a selfie evaluation from Selfie Bot is simple: simply tweet the picture to @deepselfie.
Karpathy wrote in a blog post that in order to train the computer’s program on what good and bad selfies look like, he fed it 2 million selfies along with the number of likes each received, controlling for the number of followers. After analyzing each photo “several tens of times,” the program “tuned its filters” to differentiate between good and bad selfies, according to Karpathy.
Karpathy then ran another 50,000 selfies through the program to suss out the commonalities shared among what the computer thought were good selfies. Based on the results, Karpathy writes that taking a good selfie means doing:
– Be female. Women are consistently ranked higher than men. In particular, notice that there is not a single guy in the top 100.
– Face should occupy about 1/3 of the image. Notice that the position and pose of the face is quite consistent among the top images. The face always occupies about 1/3 of the image, is slightly tilted, and is positioned in the center and at the top. Which also brings me to:
– Cut off your forehead. What’s up with that? It looks like a popular strategy, at least for women.
– Show your long hair. Notice the frequent prominence of long strands of hair running down the shoulders.
– Oversaturate the face. Notice the frequent occurrence of over-saturated lighting, which often makes the face look much more uniform and faded out. Related to that,
– Put a filter on it. Black and White photos seem to do quite well, and most of the top images seem to contain some kind of a filter that fades out the image and decreases the contrast.
– Add a border. You will notice a frequent appearance of horizontal/vertical white borders.
According to Karpathy, bad selfies identified by the program tended to have low lighting and up-close framing where the head was too large. Group shots also tended to fare poorly.
However, there is one major caveat for the program’s algorithm pointed out by Quartz: assessing selfies as “good” based on likes favors younger, lighter-skinned women, since their photos are more likely to be on the receiving end of netizens’ likes.
For those who want a selfie evaluation from Selfie Bot, the process is simple: simply tweet the picture to @deepselfie and then await its reply. Whatever the grade, don’t take it personally — it’s only a robot after all, and selfies can be improved by following good selfie tips.