Meta's AI tool won't draw without a sketch assist from you

Meta's AI tool, Make A Scene
(Image credit: Meta)

AI-generated images are all the rage right now with people using Craiyon (formerly known as DALL-E mini) to create nonsensical memes. Now Meta is throwing its hat into the ring with its own image generator, but with a bigger emphasis on creating artistic pieces. 

The AI tool is called Make-A-Scene and it uses a combination of text prompts and sketch outlines to generate a 2,048 x 2,048-pixel image. Meta demonstrated the tool in a video where the company asks several artists to create an art piece.

One artist, Alexander Reben, prompted Make-A-Scene to create a high-definition image of a hot dog floating in the sky and entered into the system a simple sketch of an oval above a square. Sure enough, Make-A-Scene created a detailed image of a floating hot dog above a body of water. The sketches themselves don’t have to be super detailed. It’s about giving the AI a general idea of how the final image is supposed to look.

AI art

Meta added sketches as part of the image generation process because, it claims, text prompts alone don’t do a very good job of reflecting an artist’s intent. The company gives the example of asking the tool to create an image of a zebra riding a bike. With only the text prompt, the zebra in the image may be too big or facing the wrong direction. Sketches give Make-A-Scene a structure to follow.

Images created with Make-A-Scene are more faithful to the sketch outline rather than the text prompt. According to developer tests and evaluators, generated images were found to be “better aligned” to the outline 99.54 percent of the time. The tool took the text prompt into account 66.3 percent of the time.

It’s unknown if and when Make-A-Scene will be made available to the public. Meta alludes to the tool being used to help people with their own creative endeavors or an art director with their marketing campaign. We also don’t know how the sketches will work.

Will people be able to upload their own outlines to the image generator or will they be provided by developers? We asked Meta for more information about the sketches and we'll update this post if it ever responds.

Meta states that it will be holding a presentation for Make-A-Scene at this year’s ECCV event (European Conference on Computer Vision) which is being held in Tel Aviv, Israel. The event starts on October 23 and goes on until the 27th.

Analysis: Future with AI

AI generator tools are a fascinating concept that has created some incredible results. We created some of our own unusual AI art with Google's Imagen AI generator. 

AI imaging tools aren't quite ready to rival skilled artists and most aren't available to the public due to fears of misuse, aside from Craiyon which uses an open source model to replicate OpenAI's DALL-E. But there are developers creating AI tools outside of art creation. Namecheap, for instance, is an AI tool that can generate a name for your business if you’re stuck on ideas.

There are also AI writing tools to help quickly compose content. In fact, we have a list of the best ones for 2022, but they do require human intervention to edit them. We just hope they don't get too good or we're out of a job.

Cesar Cadenas
Contributor

Cesar Cadenas has been writing about the tech industry for several years now specializing in consumer electronics, entertainment devices, Windows, and the gaming industry. But he’s also passionate about smartphones, GPUs, and cybersecurity. 

Read more
An image created by Google's Imagen3 artificial intelligence image generator.
What is Imagen 3: everything you need to know about Google's text-to-image model
Artbreeder
What is Artbreeder? Everything we know about the AI art platform
Google Whisk
I turned my dog into a plushie using AI and it was super easy
An AI-generated image of the city of London in the future, created by Dall-E 3 using ChatGPT
What is DALL-E 3: everything you need to know about the AI image generator
NightCafe
What is NightCafe? Everything we know about the AI image maker
The Google Gemini logo against a black background.
I tried Gemini's new AI image generation tool - here are 5 ways to get the best art from Google's upcoming Flash 2.0 built-in image upgrade
Latest in Pro
Isometric demonstrating multi-factor authentication using a mobile device.
NCSC gets influencers to sing the praises of 2FA
Sam Altman and OpenAI
OpenAI is upping its bug bounty rewards as security worries rise
Context Windows
Why are AI context windows important?
BERT
What is BERT, and why should we care?
A person holding out their hand with a digital AI symbol.
AI is booming — but are businesses seeing real impact?
A stylized depiction of a padlocked WiFi symbol sitting in the centre of an interlocking vault.
Dangerous new CoffeeLoader malware executes on your GPU to get past security tools
Latest in News
Nintendo Switch 2 Joy-Con up-close from app store
Nintendo's new app gave us another look at the Switch 2, and there's something different with the Joy-Con
cheap Nintendo Switch game deals sales
Nintendo didn't anticipate that Mario Kart 8 Deluxe was 'going to be the juggernaut' for the Nintendo Switch when it was ported to the console, according to former employees
Three angles of the Apple MacBook Air 15-inch M4 laptop above a desk
Apple MacBook Air 15-inch (M4) review roundup – should you buy Apple's new lightweight laptop?
Witchbrook
Witchbrook, the life-sim I've been waiting years for, finally has a release window and it's sooner than you think
Amazon Echo Smart Speaker
Amazon is experimenting with renaming Echo speakers to Alexa speakers, and it's about time
Shigeru Miyamoto presents Nintendo Today app
Nintendo Today smartphone app is out now on iOS and Android devices – and here's what it does