(Phys.org)—There you are passing time on a subway with a stalled train in the middle of a tunnel and no comforting announcement saying why.
You get out your phone, scroll, scroll through your most recent picture collection. Eight pictures of the same cat in the same box, 13 pictures of the same kid on the same skateboard, two pictures of the same grilled cheese sandwich, seven of some new machines in the lab and three tourism shots.
Not necessarily in any logical order relevant to your life and definitely not in any order of quality. It would be nice to have the functions that go beyond merely logging what time, day and month the picture was taken.
EyeEm is behind the app that offers a higher experience in finding photos and deciding which to share. The enabler is in the app, in the name of image recognition technology.
The recently announced iOS app is The Roll, an "intelligent" finder that can make real sense of your photo collection. "Photos are automatically tagged and displayed in categories so you can search and share without clutter." Specifically, The Roll organizes, tags and finds the best photos on your phone.
DL Cade in PetaPixel: "The premise behind 'The Roll' is quite simple: without any input from you, it will automatically organize, tag, and rate all of the photos on your iPhone. Each photo is given an "aesthetic score" between 0 and 100, multiple nearly-identical shots are bunched together with the best one on top, and tags are automatically added to your images so you can actually find them later without scrolling to infinity."
The company said that EyeEm Vision is getting smarter every day. This means you can find the photo you are looking for in a flash.
PetaPixel said the company's image recognition technology has been trained by millions of images. "You can think of that computer vision tech as the app's secret sauce."
The company site said "Trained using millions of curated photos, EyeEm Vision learns and replicates the choices of professional curators and applies scores to photos from 0 to 100. While this aesthetic taste totally depends on your personal preferences, The Roll helps you to quickly identify your best shots."
Appu Shaji, the head of research and development at EyeEm, wrote an article in February this year about their work and their approach.
He said that "Often our understanding of a picture is influenced heavily by our previous visual stimuli. A much stronger strategy is to transform the image via various mathematical operators into a set of measurements called image features that summarize the visual information that is relevant to our solution, and disregard irrelevant information. Finding a good feature representation manually is hard, but in the recent past Convolutional Neural Networks (CNN) have exhibited tremendous success at automatically determining relevant features directly from training data."
PetaPixel's Cade tried out the app and had this to report. "With only a 16GB iPhone I have to keep my Camera Roll annoyingly trimmed, but even I have 350+ images in there that I've only scrolled all the way through maybe once or twice.
"In less than a minute, The Roll had compiled all my images, bunched duplicates, scored each shot, and tagged everything. No matter how you slice it, that kind of speed is impressive."