But it comes at a price.
You’ve probably put more thought than you’d like to admit into what you would look like as a cartoon character. What would you wear? Would you be more Disney princess or Marvel superhero? The list goes on and on. Well, you can finally see your animation dreams come to life. There’s a new app that creates artificial intelligence (AI) renderings of your selfie, and the results can get seriously realistic. If you’ve seen the AI trend make its way to Instagram, you’re probably wondering what it is, and more importantly, how you can make your own. Luckily, the Lensa app is an AI image generator that does the work for you, but the computer-generated selfie comes at a price.
The thing about trends is that it can be hard to pinpoint where it’s all coming from. With the AI trend, for example, people started posting animated images of themselves with no real explanation. Thankfully, it’s not too hard to figure out — the app everyone is using is called Lensa. It started taking over social media in early December 2022 for its “Magic Avatars” tool, which is a feature that lets you create AI-generated selfies based on real photos uploaded into the app, so you can finally get a glimpse at what the animated version of yourself might look like. Sometimes the results are less-than-impressive, but most of the time the images look like they’re straight out of a Disney movie. Plus, the app can transform you into just about anything from a superhero to a fairy princess, so you can cosplay as Captain Marvel and Cinderella.
Lensa is a photo and video editing app that’s been around since 2018. Not only does the app come with helpful features like the under-eye circles removal and eyebrow thickening tools, but it also has the digital art Magic Avatar tool, which uses artificial intelligence to generate dozens of unique renderings of a subject in different scenes, outfits, hairstyles, angles, and more.
It’s a good thing the app comes with so many perks, because a subscription costs $29.99 per year. Despite the monthly price tag, there is a one week free trial that you can sign up for. Don’t be fooled, though, because you’ll still need to shell out some extra cash along the way to make your creation. If that’s not a dealbreaker for you, here’s what you’ll need to do to make your own.
To get the process started, the first thing you’ll need to do is download the Lensa app for free from the App Store and open the app. After scrolling through a few slides, you’ll see a slide the says “Try Free & Subscribe.” Make sure to toggle on “free trial enabled.” Keep in mind that after your one week free trial is over, you’ll be charged $29.99 for a subscription for the whole year, so don’t forget to cancel if you don’t want to pay more.
Next you’ll have to set up your account with an email address, and then it’s time to get to work. On the pop-up for the Magic Avatars tool, tap the “Try Now” button at the bottom of the screen. On the next page is a disclaimer that basically says some pictures will come out better than others, which is to be expected from an AI generator. If you’re still willing to take the risk, accept the terms and conditions at the bottom of the screen before tapping the Continue button.
You’ll be given a list of instructions to follow, including using no less than 10 photos and no more than 20, selecting photos that only feature one person, providing images that have a variety of angles and facial expressions, and more. Tap the circle next to the terms and conditions at the bottom of the screen, then tap the “Select 10-20 photos” button underneath. If your photos don’t meet the criteria, you will be required to pick more.
Scroll through your camera roll to find a good variety of selfies, then import your selections into Lensa. Afterwards, you’ll be asked to select your gender, then select the number of avatars you’d like to purchase. (Note to self: not all free trials are actually free.)
You can opt to buy 50 unique avatars (five variations of 10 styles) for a one-time purchase of $3.99, 100 unique avatars (10 variation of 10 styles) for $5.99, or 200 unique avatars (20 variations of 10 styles) for $7.99. These are the discounted prices for Lensa subscribers, but it doesn’t look like there’s a way to use the app without being a subscriber, so that probably means the full price options are just a ruse. Oh, and BTW, you’ll be making your purchase selection before you get to see your AI-generated avatar, so tread carefully.
The Lensa app will upload and generate your AI selfies. The process may take 20-40 minutes, but luckily you can close the app while it works its magic and be notified when your avatars are ready. When they’re finally ready, you’ll be taken to a screen with a smiley face icon in the top left side and a message that says, “Find Magic Avatars here.” Tap the icon, then tap your pack of avatars. Your collection of AI-generated selfies will then be separated into all sorts of categories, including Anime, Fairy Princess, Fantasy, Stylish, and more.
Of course, this whole process would all be for nothing if you can’t keep the images at the end. To save the photos to your Camera Roll, tap the “Save all avatars” button in the top right corner to save all the photos in the pack. If you don’t want to clutter your Photos with the duds, you can save pictures individually by tapping a photo then selecting the “Save to Photos” button underneath the avatar.
You can also tap the Share icon next to the “Save to Photos” button to share your faves on Twitter, Instagram, Snapchat, Facebook, Messages, AirDrop, Notes, and more.
According to Tech Crunch, Lensa works by utilizing a Stable Diffusion model, which means it generates images from text descriptions of images. It uses open source material from the internet, which is legal, but some professional artists are calling the way the AI works a gray area that borders on stealing. Elite Daily reached out to Lensa and Stable Diffusion for comment on the criticism of the Magic Avatars tool and using open source code to train AI, but did not hear back at the time of publication.
Stable Diffusion is an AI which was “trained” by looking at a cross section of the internet — some 2.3 billion images across the web, some of which may have been copyrighted, watermarked, or otherwise privately owned. The AI then uses those examples to generate the images that show up as your avatar. Crucially, artists don’t have the right to opt in or out of having their posted work used to train the AI, as the training datasets are collected as URLs of original images from which the data is gathered and the picture is then discarded.
All of this is to say the open source code is legal, but artists don’t really have a say in how it’s used. Artists on Twitter like Meg Rae point out what she calls “a legal loophole to squeeze out artists from the process” to avoid paying licensing fees. Jon Lam, senior storyboard artist at RIOT Games, also pushed back on the trend for similar reasons and added that there isn’t enough knowledge about the topic. He wrote in a Dec. 5 Instagram Story, “Artists, people care, they just don’t know how AI is ripping us all off.”
Editor’s Note: This post was updated to include further information about the Stable Diffusion model and how it works to create AI images.
Be the first to know what’s trending, straight from Elite Daily
This article was originally published on
source
—
Note that any programming tips and code writing requires some knowledge of computer programming. Please, be careful if you do not know what you are doing…