We're not making this up.
Think you got the selfie thing down? Think again.
A computer science graduate student at Stanford University found the secret to the perfect selfie. A Huffington Post article narrates how Andre Karpathy "trained a Convolutional Neural Network, a type of data mining network capable of processing 140 million different specifications, to judge whether a selfie was successful or not."
To save you the nitty gritty bits of how he went about gathering this data (which you can read about on his blog), here's a breakdown.
Karpathy collected two million photos with the hashtag "selfie." He then analyzed the number of likes per follower, and ranked the ones with the most as "good selfies" and the rest with a "negative" label. Then, he sorted those photos by the number of follows.
"I gave a small bonus for each additional tag on the image, assuming that extra tags bring more eyes. Then I marched down this sorted list in groups of 100, and sorted those 100 selfies based on their number of likes. I only used selfies that were online for more than a month to ensure a near-stable like count," he said on his blog.
This is what he found, and it can help you create the perfect selfie:
- Be a woman. He found no men in the top 100 photos.
- Use the rule of thirds. "The face always occupies about one-third of the image, is slightly tilted, and is positioned in the center and at the top," he noticed.
- Have long hair.
- Use filters.
- Add a border.
Take a lesson from Kim Kardashian, for example:
All photos: Instagram
Are you a man and devastated that your selfie didn't make it to this Top 100? It's OK! Karpathy notes that the rules are different for men and celebrities. Now, show the world your best selfie.