The secrets behind how your fave Snapchat filters work

The secrets behind how your fave Snapchat filters work

It's doggone amazing!

In those moments of boredom when you're playing with Snapchat's filters - sticking your tongue out, ghoulifying your features, and working out how to get the flower crown to fit exactly on your head - surely you've had a moment where you've wondered what's going on, on a technical level - how Snapchat manages to match your face to the animations?

Vox has put together a concise, informative video explaining how Snapchat's filter technology (or 'lenses', as they're referred to at Snapchat) work. In the video, it's explained how Snapchat acquired the software off a Ukrainian company called Looksery in 2015 for $150 million. The Looksery technology is similar to the technology Facebook uses when it suggests tags for your photos - it collects pixelation data from the contrasting light and dark areas of your face (when it's head-on) - which allows it to determine its position. The filter then locates the features, and a computer vision 3D mask then acts as coordinates to trace your movements.

Just last month, Snapchat acquired another software company - 3D photo app maker Seene, which lets you capture 3D models from your phone's camera. Snapchat will possibly use Seene to create even more selfie lenses - in a new 3D photo format. Seene’s team, which is based in the UK, is relocating to Los Angeles, where Snapchat is headquartered. Watch Vox's video below:

Related:

Pilerats' Guide to Snapchat Users

To Cull Or Not To Cull Pt. 2

Colby Bignell from Perth punkers Flowermouth offers his (anti-killing) thoughts on the matter.

4 years ago

No Money Weekend June 26

Your essential nation-wide cheapskates event guide.

2 years ago

The Bell Tower Times Guide To Bar Etiquette

Some hot tips for being hot shit at the bar this weekend.

2 days ago

Get around Cleverman - the best new Aussie show in...forever?

Replacing and removing racial stigmas and stereotypes one great piece of story telling at a time.

1 year ago

Close
-->