Online photography platform and community EyeEm has unveiled a new feature designed to automatically unearth the best photos on your smartphone.

With EyeEm Selects, mobile photographers who have EyeEm installed can elect to have their camera roll scanned by the app, which then uses computer vision to suggest the best photos to upload to the EyeEm community.

Founded out of Germany in 2011, EyeEm, in its purest form, is a 20-million strong community of photography enthusiasts who use the platform to upload photos, share tips, and discuss all things photography. It has also evolved into a marketplace for buying and selling images and recently expanded into videography, emulating photo-centric platforms like Instagram, which also began life in the still-image realm.

Computer vision-ary

Back in 2015, the Berlin-based company unveiled EyeEm Vision, which was a computer vision-powered engine that automatically tags photos based on what it detects in the image. And EyeEm Selects is basically built on top of that technology. It’s also worth noting here that the latest feature doesn’t rely on server power to process the imagery — everything happens within the app on your device, which goes some way toward not only preserving battery and bandwidth but also ensuring privacy.

The problem that EyeEm is looking to fix here is one that’s familiar to millions of people around the world. With the advent of smartphones, most people have a decent-quality camera in their pocket, which means they end up taking hundreds or thousands of photos, many of which may never see the light of day.

And with EyeEm Selects, the company is not only looking to encourage users to download and install its app to help curate their snapshots, it’s also hoping people will be tempted to upload their photos to the EyeEm platform.

Above: EyeEm Selects: Scanning

The concept behind EyeEm Selects isn’t new. Google, for example, offers a similar feature within its Photos app, while Israeli startup MyRoll has an app called Gallery Doctor that uses algorithms to free your smartphone from bad photos. However, the Gallery Doctor app hasn’t had many updates since AVG acquired MyRoll back in 2015. Elsewhere, fellow online photo platform Shutterstock is now using machine learning technology to allow users to carry out reverse image searches for stock photos.

In terms of the criteria EyeEm uses to determine a “good” photo, the company said it strives to surface images that “…communicate strong stories with good composition, and [are] shot with technical mastery,” according to a company spokesperson.