Taking photographs of yourself, your family, you friends, your vacations, your food, and a host of other things is possible because you don’t have to pay for film anymore. We all know this. With the explosion of digital memories and documentation of almost every aspect of your life, you now have a lot of information tucked away, and many artificial intelligence applications are aiming at identifying what is in them, helping you manage them, and modifying them, among other things.
You can expect more and more coming from AI in the coming months, but here is a short list of what can currently be done with your personal photographs.
Machine Learning Can Identify What Is in Your Photographs
1. Identify People, Objects, Scenes, and Emotions
Artificial intelligence is getting better at identify what is in your photos, and this manifests itself in at least five but no more than 10,000 useful ways. For example, you could use an app like CamFind to snap a photo of that succulent Jade plant you have resting near your south-facing window and have it tell you that you just took a picture of a “Green leaf plant in brown pot.” Just what I needed—confirmation from my AI photo assistant that my plant is in fact a plant, in a brown pot.
Trivialness aside, these seemingly simplistic classifications can be used in some very useful ways. Throw some of your photos into Google Photos, and you’ll start to see what AI can really do. It can identify people’s faces and automatically groups photos based on the people in them, it can identify places and events (from specific places like “Milan, Italy” to types of events like “concert”), which allows you to search your photos effortlessly. Type in “Buddha in Thailand,” and you instantly get back all 86 pictures you took of the reclining Buddha two years ago while visiting Southeast Asia.
Pro Tip: It can even find emotions. When I type “angry” into my Google photos, I get the bust of some forgotten historical figure I snapped a picture of in an art museum last November who looks particularly stern.
2. Live Narrate the World around You:
Microsoft has an smartphone app, Seeing AI, that aims to narrate the world around you. You can point it at a document, and it will read it to you, you can point it at a friend, and it will tell you who they are whether they are feeling sad, happy, or surreptitious. It will identify products for you if the barcode is in view, recognize the amount of money you have in your hand, describe scenes, identify colors, and read handwriting. It’s not perfect, but it’s probably the closest you’ll get to a AI assistant that narrates the world around you.
3. Identify if Your Baby Has Eye Cancer
What if you found out your infant had eye cancer? That news would rock anyone’s world. But what if you had a tool that helped you catch it early enough that your baby didn’t have to lose his or her eye and didn’t have to go through chemo? You’d probably do almost anything to get it. There’s a machine learning algorithm that can detect cancer that no human eye can detect.
It’s a terrible disease, and if it’s noticed too late it can have terrible consequences. But what if you had a way of catching the cancer early enough that it could be taken care of easily without any harsh treatments like chemotherapy?
A team at Baylor University, lead by Bryan Shaw, whose son went through the harrowing experience of eye cancer, has created an app that can detect this cancer in the eyes of regular digital photographs. This is one you will definitely want to add to hopefully catch eye cancer before it’s too late. You can head over here to listen to how Bryan created the app in his own words.
4. Count Your Calories
Disciplining oneself to eat healthy amidst an ocean of enticing yet detrimental food options on all sides is the golden pot at the end of the proverbial rainbow. Ever present. Nigh impossible. Never fear, AI will (try) and save us from ourselves.
Foodsnap brings the promise that counting calories could be almost as easy as snapping a picture. While this isn’t a problem machine learning can solve on its own with current technology (have a listen to an interview we did with Hilary Mason who discussed the challenges), it can at least make it easier.
Take a picture of your food, and the AI will try and identify all of the food items on your plate. Combine that with a sleek user interface, and the app helps guide you with as little friction as possible to a reasonable estimate for the calories in your food.
AI Can Modify Your Photographs in Surprising Ways
5. Turn Your Frown into a Smile
This one isn’t available yet to the common user, but it’s a good example of what machine learning is capable of and what is coming in the future—the ability to alter your face or other aspects of your pictures as though they were originally part of the picture.
This case is simple. Frowning in the picture? Let’s make you look happier. AI can turn your frown into a smile.
6. Stylize Them a la Van Gogh
Prisma is one app floating around that can take any of your mundane, seemingly unimportant pictures and turn them into Florentine masterpieces. With several different “styles” to choose from, you can turn your photo of your lawn gnome into a stirring piece that looks like it came straight from the Baroque, or a vibrant and warm shot as though it was taken on the beaches of Honolulu.
7. Generate a 3-D Image of Your Head
Here is a little gem from the University of Nottingham. Upload a picture of your face (or someone else’s, or Einstein’s) and their Volumetric CNN Regression algorithm will make it 3-D. It may not be endless fun, but you can’t deny the novelty.
8. Soup up Selfies
Maybe you can’t quite get the selfie perfect. No problem. You can modify how you look after the fact. Adobe’s Sensei tool uses machine learning to give you a powerful set of features you wouldn’t have otherwise. You can adjust the depth of field, blur the background, tilt the angle of your face, and reproduce your photo with another photo’s style.
Machine Learning Can Manage Your Photos
9. Deleting Duplicates and Blurry Pictures
Today photos are free, and so we take hundreds of them. Aside from the small limitation of the memory of your device, you can snap photos all day long, and it won’t cost you anything. So instead of making sure you get the perfect shot when you take it, now the problem becomes shuffling through all the riff-raff and selecting your best photos after the fact. And deleting the bad ones to conserve your memory.
AI to the rescue. Zyl is one company with an app that will look through your photos and identify duplicates and get rid of the ones that are more blurry, keeping only your best shot. Saves you time, saves you memory, and makes your photo gallery look sophisticated.
AI Can Do Other Random Things
10. Tell You How “Hot” You Are
This is probably the most vain AI we’ve come across. It takes a picture of you and gives you a “hotness” score. Are you hot? Are you not? Let an artificial intelligence be the judge. That seems like a good idea.
Disclaimer from their website: “It is important for a person to take the Facial Attractiveness Test in a fun and lighthearted manner.”
Bonus—Find Your Fine Art Look-alike
This one has been all over the social media circles by now. Snap a picture of yourself with Google’s Art & Culture app, and it will tell you who your fine-art look-alike is. Although, it’s not available for everyone.
This is another great example of an AI that can both come up with some shockingly good results, and incredibly bad ones. You could be one of the lucky ones Google finds a good match for. You’ll never know until you try.
More to Come
The capabilities that machine learning is bringing to the photography space will continue to surprise us. If you come across anything interesting, let us know about it! If you want to stay up on the newest and most interesting applications, we send out a newsletter of a few of the most interesting applications we see each week. It’s a fun read.