Simple AI-Enhanced Workflows for Video Creators – Demonstrated Live and Easy to Implement

by Mascha DeikovaMarch 22^nd, 2024

Artificial intelligence is a topic of considerable controversy nowadays, causing intense discussions everywhere. We know because we write a lot about AI tools for filmmakers and we read the comments. There are both fears and high expectations surrounding it. And yes, this topic also casts light on a big ethical dilemma, especially considering generative AI. However, if you decide to turn your back on state-of-the-art technology completely, you may miss out on some useful tools that can simplify and accelerate your work by far. In the video below, I demonstrate a variety of simple yet powerful AI-enhanced workflows for video creators that are possible to implement straight away.

For your information: this is a live, recorded session from Videocamp in Vienna, Austria. Originally, it was supposed to be just a brief introduction to the vast world of advanced technology and showcase some practical AI tools for video creators. However, a week before the event, OpenAI announced their text-to-video generator Sora, which took the world’s attention by storm. So, we decided it was also important to talk about generative AI in general, and the ethical topics surrounding it at the beginning.

The second part of the video is dedicated solely to simple workflows based on deep-learning models, which are quite helpful in pre- and post-production.  

Why should you keep up with AI development?

Although a lot of people despise AI and would rather ignore it, there are solid reasons why we believe you shouldn’t. First, the tech is already out there and is developing at breakneck speed, getting crazier every week. We can’t put this genie back into the bottle, so we have to learn how to deal with it.

Secondly, artificial intelligence is not completely evil (in case that’s what you think about image or video generators that scrape the entire Internet and train on other people’s work without credit to the original artists). There are AI upscalers that allow you to increase the quality of archive footage, DaVinci Resolve’s Magic Mask, automated captions in editing software, and much more to explore. If you don’t learn how to use these tools, at some point you will for sure become much slower than those who do. Also, if deep-learning mechanisms can take over mundane tasks leaving you more time for creative endeavors, why shouldn’t they?

In the video above, we talk about a few more reasons why you should keep on educating yourself on this topic. One way to do so is to read our articles from the “Artificial Intelligence” section.

AI-enhanced workflows for video pre-visualizations

One of the tools I showcase in the live talk in detail is Luma AI. This is a zero-cost app (also available in the web browser) that turns a simple video into a 3D model or a 3D environment using the so-called NeRF technology. Personally, I use it during location scouting and for video pre-visualizations. Even if you don’t have the newest smartphone (mine is a dying iPhone X), you will be able to share a comprehensive 360-degree look at a chosen space with your DP or production designer within minutes. Better yet, by animating the keyframes directly in the app, anyone will be able to pre-visualize exact shots on location (you can even include your characters or important props) after you’ve closed it.

Some online creators also use Luma AI to create impossible shots and cool visual effects using only their phones and imagination. Here’s an example by a vlogger @karenxcheng to give you a taste of it:

Generative AI and how it evolves

In my opinion, it is also totally okay to use generative AI in the pre-production stage: to create artistic mood boards with Midjourney, to research overseas locations with ChatGPT, or to make pre-visualizations of your desired shots using AI video generators (Runway is the one we’re taking a closer look at in the video above).

Over the last year, we’ve seen how quickly this technology has evolved. For example, here are the results created by the image generator Midjourney using the same text description: its very first model (on the left) and the actual one (V6 on the right).

No wonder this rapid advancement is so intimidating. That’s why when OpenAI’s Sora came out, a lot of video creators instantly felt they were going to lose their jobs in the not-so-distant future. Why? Because it showed unprecedented realism. We don’t dive into Sora in the video because it was the topic of someone else’s presentation, but if you want to learn more about it, head over here. Also, please, join our open discussion there, as we believe Sora is not the end of human filmmaking and would like to read your take on it.

How to enhance audio with AI?

As you know, sound is a drastically important part of every video, even if it is only an Instagram reel. However, as an independent video creator, you don’t always have the best possible conditions to record sound. Sometimes, there is disturbing irregular background noise; other times human mistakes occur and the audio has not been leveled correctly. In the past, such things were difficult to fix. Nowadays, we have AI tools to take care of it.

In our video, you will see two workflows. One helps to enhance the quality of audio within seconds. Another can almost seamlessly separate voice and music (or sounds) from each other. Let’s say you have an old recording, the project file is lost, and you need only the interview part from the audio track. That’s where such tools can be of immense help.

Other AI-enhanced workflows for video you will learn about

Last but not least, we take a look at AI-enhanced post-production workflows. In particular, you will get a step-by-step guide on how to quickly translate your subtitles into different languages using automated captions and ChatGPT. The workflow suggested in our video still requires a couple of actions to be done manually. Yet, if you often work with long interviews or documentaries, you need to try it out because it will save you a massive amount of time!

Conclusion

If you haven’t worked with AI tools yet, this video presentation might be a lot to digest. At the same time, it’s only a tiny fraction of what artificial technology already offers filmmakers and video creators. We believe it’s important to follow this topic and we will always try to keep you updated here on CineD.com. So stay tuned!

What about you? Do you already use AI? Can you share other simple AI-enhanced workflows that you find particularly practical? Let’s expand our toolkits together and talk in the comments below!