Artificial Intelligence comes to animation and graphics, part 2

Robots are already here. No, don’t look around — you are actually using them for a while now. Maybe the expectations on how robots should look mislead you, and have not found any of the classic mechanical creatures, you failed to see them at all. The best robots today are just web pages in your browser that can do something sophisticated, and, for many people, quite daunting.

With your command, of course. Without it, any robot won’t put millions of people in a state of excitement. We are back to talk about Artificial intelligence once again and measure its impact on our beloved animation and graphics. There is a lot of it. The way it is developing is surely a speedy one, and its pace is enormously quick. The question that many people might have at the moment is the following: are animated studios going to use it in their work?

While the answer has rather a variety of dimensions, handing it to you is our primary task.

Hiring AI to work for you

Let’s get a little bit boring, and take a look in the dictionary. What do we see here? “Artificial intelligence leverages computers and machines to mimic the problem-solving and decision-making capabilities of the human mind”. Hmmmm. Even though it sounds scary, the keyword for us is this “mimic”.

To replace an animated studio is impossible. We are looking at AI as a tool that helps to redirect our creative energy, come up with new ideas, or just observe this fella. It is hard to call it co-creation, because the work of AI in this respect is to provide us with research and inspiration. And, what is even more important, is to use it as a calculating machine. So, in a way, you hire AI to do the job it is really good at. But even for enthusiasts of the technology, it is quite hard to imagine AI is doing so well that it can replace a whole studio of people with different, and often profound skills in the animation industry. Is it capable of executing the story? Nuh.

Jumpstart magazine has described the role of AI in animation in the following way by taking as an example how technology is used by Walt Disney studio:  “In 2018, Disney formed the Direct-to-Consumer & International Organization (DTCI). It aims to use ML and deep learning (a type of AI and ML that teaches computers to acquire knowledge like humans do—through examples) tools to automate metadata tagging and speed up the archiving process, where people input different data into a centralized database for easy reference.

Together with Amazon Web Services (AWS), they have created an archive called Content Genome that can automate the digital archival of all Disney content. This archive can help animators find a specific reference shot or sequence from all the Disney animated films or concept arts. The new archive from Disney uses AI and ML to automate the tagging process so that animators do not need to manually tag information themselves. This program can automatically tag the content into the database to speed up the tagging process. The tags, which contain descriptions of animated films, can identify different characters in a film, their relationships, the film soundtrack, and the background scenery. As a result, writers and animators can now quickly search for and familiarize themselves with the sources without manually going through every film. This also reduces the amount of work required to organize the Disney library since people just need to approve the algorithms’ tags instead of manually creating tags by themselves”.

That’s the spirit. That’s actually what we need from AI: to help us make quicker some parts of our work that we can’t do so quickly. We’re not some robots, after all.

On testing hypotheses

Modern AI requires too much from us humans. To call it independent is ridiculous, simply because a person has to plan and describe scenes while an AI generator can only, well, generate things.

Of course, there are things like deep nostalgia that help us to see how, for example, Mona Lisa would smile. But is it really going to affect the work of a studio? Nope. It is reframing our views and helps to understand better what people are amazed by.

AI comes into more and more creative projects, and it is obvious to anyone that soon creating something without AI would be anachronistic. What we – and many other studios – found very helpful about AI is its ability to provide the quick test of hypotheses. What does it mean? It is no secret that to hack growth, many companies now use an approach where they should test new ideas before implementing them.

The old times when a company could mindlessly launch a product without actually checking with an audience are already behind us. After finding the right hypothesis, and validating that choice, they start working on a project. In the case of an animation studio, a company often has to present a mood board for a client in order to give a feeling and factual information about the project. You can imagine that brainstorms take time, and creating a comprehensive mood board in minutes is impossible. But it is possible to work with AI in a way that helps move really quickly.

Dozens of times quicker. However, the results are not always satisfying, but it helps us move. It is also allowing us to understand better how to shape what we want regarding a certain task. So, AI will give you a lot of support when creating a mood board, but it is helpless when it comes to narrative challenges: only your scriptwriters can really make it work. This is one of the weaknesses we find in AI. Let’s see what else they are not good at.

Weaknesses of AI generators

Yes, these robots are not that perfect. Not always you get high resolutions, not always a piece is properly finished. If you want to revise it properly, well, you can’t. It seems that a number one priority of these things is entertainment, even though artists use it as “their” artwork and actually convince people of that. Watch part 1 of this article.

If you type twice the same description for an image, you will get different ones. It is not necessarily bad, but it only makes us think that working seriously with it rather too soon. We admire the power of machine learning. However, pre-defined sets of animations basically defines the whole technology behind AI. This is raising a fair question: is it even possible to use AI to create animations? This is another weakness: only after “feeding” an AI generator a bunch of images, it can make something saucy out of it.

One of the trends of AI animation is reconstructing 3D human motion from 2D content. The synergy of computer vision, deep learning and anthropometry allow achieving great results.

Other uses of AI in animation

The gaming industry is using animation more than any other field. It includes training digital characters to perform complex motor skills right – in a human-like, convincing manner.  Walking, dancing, athletics, martial arts are AI taught, so to say. The movement is generally something that AI is good at providing. Designing not only walking, but also talking of the character is now the task for machine learning technology.

The animated character’s mouth movement and its synchronicity with the voice-over is a job for an AI. Facial animation is easier to do with this technology, and many projects are using it. When you work with the animated projects, you often accumulate much of the raw material that will become the final result at the end. AI might provide you with the images and decisions that might give you an anchor for a future of the project.

Summary

Regarding an animated studio, AI can be really helpful at speeding up the production process, therefore, saving time for team members to do something more valuable. AI is a worker that won’t get tired, and may easily provide you with vital information to finish a project or make its process quicker. For those who are afraid that in a few years AI will make us quit and be able to do all the things on its own, we can say only one thing: don’t. Don’t be afraid of that. Humans evolve by their own curve, and AI only boosts our evolution.

If you are using AI smartly, you can learn from it and improve your own work. Not only in animation. By using AI to improve your own creative decisions, your company will win, bringing clients out-of-the-box ideas. The imagery that AI offers to all creative people shouldn’t be perceived as a part of the competitive narrative. After all, we create those things to help us – not to overcome our abilities. Write in the comments how you see AI and don’t be afraid to share your views with us! We are always appreciating how you see the things we write about here.