“I visualize a time when we will be to robots what dogs are to humans, and I’m rooting for the machines.”
Claude Shannon
Animation is an incredibly unique form of art. It can be very personal and intimate, but it can also be incredibly vast and open-ended. AI is rapidly changing the world of animation, and it has the potential to be even more transformative than ever before.
Let’s take a look at some of the ways that AI is changing the animation industry today and how it could shape the future of animation as a whole.
The above introduction was written by a bot called Jenni. Jenni is an Artificial Intelligence (or AI) writing assistant that works with keywords and phrases to assist you in writing blogs quicker and more efficiently. If AI can take industries like writing, what about animation?
In this blog post, we’re going to do a deep dive into AI animation and whether or not it really can be the future of animation. Unfortunately, Jenni will be sitting this one out for now, but we’re glad she provided a little bit of help. Thanks, Jenni!
What Is AI In Terms Of Animation?
Considering what we’ve seen above in terms of an AI writing effectively and like how a human might actually speak and write, think now about how many blogs you might have read in the past that might actually have been written entirely by AI.
How often do you think an online piece of writing has been written and uploaded by a robot, with a human behind the editing stage of the process? Who can accurately say?
Now consider how soon it might very well be that AI takes over the animation industry as a whole. Thankfully, it might not be as soon as you think.
When we think of an art form such as animation, we immediately think of creativity, freedom, and expression above all else. Of course, we might also think about movies and series or shorts created by artists for entertainment, but as we’ve seen over the years, there are many other purposes animation has.
For those unfamiliar with the concept of AI: AI is essentially a simulation of human intelligence in machines - in most cases computers - that are capable of learning and improving their actions based on their experiences, i.e., with whatever is fed to them.
In terms of animation, it’s more of an aid for automation and 3D modeling, but there are a few other uses currently in circulation.
So let’s look at how AI was explicitly brought into the animation industry and how it’s affected the industry as well. We’ll talk about the history, as well as some examples of its use, such as in Rotoscoping and 3D Face Modeling.
Innovation In Animation
In general, we can attribute much of the innovation in the animation industry to Disney Studios, their animators, and their want and need to experiment and broaden the horizons of what is thought possible with animation.
The example of Disney utilizing Rotoscoping heavily in their early works, such as Snow White And The Seven Dwarves, is one such example. They might not have invented the process - we have Max Fleischer to thank for that - but they certainly made stern use of it in their animated feature films.
Rotoscoping is something we talk about in detail in another post, but the simplest explanation is this: it’s a process of tracing footage frame by frame onto glass to produce realistic action in their scenes.
Animation techniques like Rotoscoping have now been effectively automated by AI software like Comixify. It’s incredible just how quickly solutions to problems can be found with extensive amounts of research.
Working our way through the years, we see the use of Deep Canvas working simultaneously with the CAPS system (or Computer Animation Production System) developed by Disney to allow animators to create entire 3D backgrounds and environments for their characters to live in.
And now, we’re facing a surge of AI and real-time animation in the world of animation, and with the ever-increasing demand for high-quality animations created in as short an amount of time as possible, it only seems natural that we’d turn to incorporate machine-learning software to help.
Yeah, okay, but exactly does AI animation mean?
The Introduction And Induction Of AI Animation
The above gif, as well as the one before that, both belong to a music video made by Intel in partnership with Chinese singer Chris Lee (Li Yuchun). We highly recommend you watch the entire video, even if the music isn’t right up your alley, simply because of what it represents.
The AI in question might very well not be as impressive as you think, but the mere thought of a fully AI-driven music video is astounding. The fact that she didn’t need to wear any tracking markers on her face - a hallmark of traditional animation tracking - is the truly incredible part.
Intel instead took still photos of her face and set the AI to produce a 3D mask, allowing the animated effects to be overlaid in real-time.
Intel’s systems can assist in drastically lowering both production time and costs, which means a heck of a lot for both small and large-scale animation studios around the world.
The history of animation that includes a connection to the world of science is not something to be ignored. In the early days of Disney’s animation studios, animators were required to take anatomy drawing classes to comprehend the mechanics of human movement fully.
And so it makes sense that, in the digital age, we see a combination and conglomeration between art and science once more.
We first saw instances of Computer-enhanced AI in the 1950s, but we are leagues ahead in technology thanks to advances in both hardware and software, and our machines running on AI are now well capable of analyzing images, colors, pieces of data, and many more, to create incredibly impressive computer-generated pieces of art.
Image via Unimersiv
Dr. Llyr Ap Cenydd is responsible for having worked on the first Virtual Reality exclusive aquatic safari park game, Ocean Rift. The game, which came out in 2019, is one of the classic examples of AI and science combined with art and biology to create virtual life.
The Different Uses For AI Animation
Image by ABC Network via Giphy
Rotoscoping
We mentioned Rotoscoping earlier, a rather laborious process nowadays that finds itself being handed off to AI and deep learning software to automate the process fully, and the time it takes to finish the process, as well as how much it costs studios, has vastly been decreased.
Image by Shane O’Neill via Looper
You may be familiar with the 2009 animated film Coraline. LAIKA, the company responsible for Coraline and many other fantastic titles, has partnered with the technology titan Intel too; according to Ritika Sagar, “combine machine learning and Artificial Intelligence to develop tools that will accelerate rotoscope tasks.”
Their films use a unique 3D-scanned facial animation that has since started using recognition software to automate and smooth out the process. AI is fantastic at making repetitive tasks simple and not so repetitive, and Intel is focusing on making tools that assist in streamlining the entire design process, something that often takes big studios years.
Machine learning requires large amounts of data to analyze and work off of, but Intel and LAIKA decided instead to work with smaller amounts of data for their stop-motion animation process.
3D Facial Modelling
Gif by Disney Research via AnalyticsIndiaMag
Objectively, the process of animating faces is fairly simple once you get the hang of it. But, animating your face to be realistic? Now that’s tough.
Perfecting the subtleties of human emotion and expression is an incredibly taxing task and one that can take ages to get right.
The conventional methods involve taking models that are built from 3D face databases that use multi-linear morphable models, but given its linear nature, you lose the expressivity of the model. Linear models also lend themselves to creating impossible facial expressions, which is where AI comes in.
Disney researchers have taken to adopting deep learning neural networks, with a focus on generating realistic 2D faces.
They have suggested utilizing neural architectures to combine both linear and non-linear methods of face modeling to provide semantic control over both aspects. It uses various techniques like 3D face synthesis, facial performance transfer, performance editing, and 2D landmark performance retargeting.
The software doesn’t explicitly learn to disentangle the identity and expression of the model but rather imposes it directly onto the architecture.
But we won’t be going into more detail about that right now.
All you need to know is this: the software will be able to generate a simple human face and corresponding expressions, capture a real-life human face and transfer it to a 3D character, and allow animators to edit keyframes and performances.
Voice-over
Image via Adobe
I think we can all agree that voice-over is one of the most time-consuming parts of the post-production animation process, but thanks to Adobe, this process is now simple and smooth.
Adobe has released their Sensei AI Animation Technology within their Character Animator software that uses AI-powered lip-synching to assist in matching dialogue with animated characters. Adobe’s new software allows one to assign mouth sounds to mouth shapes, and the software accurately syncs both the character movement and the audio frame-by-frame.
The Pros And Cons Of AI Animation
With a grand idea of what AI actually means in animation, let's have a look at what it means for animation. Pros and cons, here we go.
The Pros Of AI Animation
- Innovation Is King
We as a species have to move forward with technology. Not only is it simply in our nature to want to improve, but it is increasingly becoming more and more important that we figure out how to make our lives simpler, easier, and safer.
AI is simply the next step forward for the animation industry for animated works to be made both smoother and quicker.
If you don’t believe us, here’s an example: AI technologies have been introduced to assist doctors in predicting breast cancer in women at earlier stages.
- Improved Production Process Speeds
Utilizing AI technologies within other animation software technologies allows machines to make decisions faster than we ever could. Decision-making for a machine is logical and unemotional, as it delivers results based on the information fed to it and nothing else.
Think of it this way: would you rather harvest your 100-acre farmland with a scythe or with a combined harvester? If you said the former, you’re lying because the fact of the matter is that technology and machinery have saved us humongous amounts of time and effort over the years, and it’s only getting better.
- Reduction In Human Error
Humans make mistakes; this much is and will always be true. In the case of computers, however, they don’t. At least when we don’t input mistakes.
In the case of AI, decision-making is born from experience. Previously gathered information and data are compiled and analyzed, applying algorithms to make decisions.
This means that the likelihood of error-making is drastically reduced.
- 24/7 Availability
Let’s face facts here: machines don’t sleep. Machines also don’t get bored or distracted. Humans do, and that’s life. Humans function in a way that requires them to take breaks in between huge bouts of work. Humans also need to take time away for personal reasons; something machines don’t need to do.
We see the practical application of this in educational institutions and helpline bots that receive queries outside of normal working hours.
The Cons Of AI Animation
- High Costs
As with all technological innovations, the cost of implementation is high and often increases at a rate directly proportionate to the rate of innovation.
As we progress in technology, the hardware and software required to handle our demands must match. Machines also require regular maintenance and repair, and if they already cost a great deal to create, you can only imagine how much it will cost to keep them up to date and fixed regularly.
- Stifled Creativity
One of the biggest problems with AI is its inability to think outside of the box. Our ability to be creative is what makes humans so incredible and is exactly what makes AI so linear. And, of course, this might very well work for many systems, but not exactly for art forms.
Here’s an example: the bot Quill is used by Forbes to write their earning reports. While it is impressive how effectively the bot can produce articles on its own, given nothing but data and facts, it lacks any sort of unique or creative spin on its style of writing.
- Fear Of Unemployment
As is to be expected, the introduction of newer technology brings with it a fear of unemployment, and as creatives, it is our worst fear to find ourselves without an outlet.
As AI replaces the mundane and repetitive tasks often taken on by individuals with minimum qualifications, those same individuals can find themselves out of work until they find some more qualifications.
Every corporation and organization wants to get the most out of their company, and if that means replacing human effort with robots that do their job more efficiently, then that’s unfortunately what happens.
You need only look to the automotive industry for proof.
- Animation Could Lose Its Connection To Art
Animation is art. Art is emotion, creativity, and expression, all the qualities that robots don’t have.
You might remember the movie I, Robot featuring Will Smith, and more specifically, a certAIn scene where his character is questioning a robot and debunking their ability to feel emotions.
The quote is, “Can a robot write a symphony? Can a robot turn a canvas into a beautiful masterpiece?”. The rest of that quote ends with the robot responding, “Can you?” but we’ll ignore that for now.
Essentially, what AI lacks is empathy and an understanding of what exactly it takes for a human to connect with a piece of art emotionally, whether it be a piece of music or an animated film. Working in a team is also an essential part of working as an animator in a studio, something robots can’t do.
The final thought on this matter is that (for now) most machines can only perform the tasks given to them, which disallows any creative, out-of-the-box thinking that you get when you ask a human to perform a task.
Is AI Animation Going To Take Away Your Job?
Gif by The Tonight Show Starring Jimmy Fallon via Giphy
The short answer is no. Don’t worry. You won’t be losing your job to the machines any time soon.
What AI is going to do, for now, is make your job just a little bit easier, which is really what technology has always been there to do.
The long answer, however, is both yes and no. How diplomatic.
See, machine learning and AI is definitely going to replace some of the more mundane tasks currently operated by humans, such as customer service, data entry and research, and many more, but will at the same time open up the doors to other jobs such as maintenance and developmental research.
In essence, any job that can be learned by feeding a machine ridiculous amounts of information can be automated. This can seem scary conceptually, and the disruption of the job market will be akin to that of what farmers faced in the industrial revolution. And much like other technology, the rate at which AI will improve and be integrated into the workplace will only increase.
It’s not strictly a bad thing, though. For creatives, at least, the emotive reaction one gets from producing and experiencing art in whatever form is related to the human behind the art. So while some parts of the animation production pipeline might be replaced by AI software, that doesn’t mean you’re going to lose your job. It might very well make it more enjoyable.
Think of it this way: the more technology we have, the more we have to understand and deal with, and the more people we will need to help out. More stuff equals more jobs. Kind of.
So AI Animation The Future?
AI is potentially one of the most equally exciting and terrifying concepts turned realities of recent years, allowing mundane and time-consuming tasks to be automated and simplified, giving animators the focus to work on more important parts of the production pipeline.
The induction of AI into animation has introduced the ability for animators to spend more time making sure their work is of higher quality in a much shorter time.
Whatever your thought on AI or machine learning, the future certainly looks interesting.
For more info about AI Animation, the future of animation in general, as well as answers to any other questions you might have, be sure to follow our blogs, check out our free masterclass, download a copy of our free marketing handbook, and check out our blog on “How to Start an Animation Studio”!
