New AI Tool: Hedra makes your characters talk and sing for free
Hedra improves Facial Animation with AI
Hedra improves Facial Animation with AI
Another day, another amazing AI tool has emerged: Hedra.
This platform allows users to animate faces effortlessly from audio or text, making it possible to create dynamic and expressive videos from static images. Whether you want to make an image talk, sing, or display a range of emotions, Hedra will do it.
Currently, the tool is in Beta, and it’s open for free. This may change in the future.
How Hedra Works
Hedra has a great UI, it’s easy to use it. Here’s a step-by-step guide to creating your own animated content:
Define Audio:
Generate Audio: Users can type or paste text and choose from a variety of preset voices.
Import Audio: If you have an audio file, you can upload it directly to the platform. For more options, go to https://elevenlabs.io/, write your text, and pick one of the voices for free.
Define your Character:
Upload your Image: Select the portrait you want to animate, and upload it to the platform.
Describe your character: Provide a description of your character. Hedra can generate a realistic and expressive 3D model.
One possible workflow
To test the tool, I had the idea of a Mafia conversation where one of the participants is accused of being a traitor. I asked ChatGPT for a rough draft, added my own touch to keep each character different.
Then, I generated the portraits in Midjourney, using the same sref 3211368754 for the three portraits to keep the same style:
And that’s it. Take a look at the final result:
Betrayal, a mob story | 100% AI
Impressive Consistency and Versatility
One of the standout features of Hedra is the consistency of the characters it generates. The model ensures that the animated figures maintain their identity and expressive traits throughout the video, providing a high level of realism.
Human Expressions: Hedra’s model can animate characters to sing, act, and display various emotions.
Inanimate Objects: The tool isn’t limited to human faces. It can also animate expressive inanimate objects, provided they have a clear mouth, nose, and eyes.
While Hedra offers impressive capabilities, it’s not without its limitations. The model works best with images that have clear facial features, such as a mouth, nose, and eyes. And currenty, it only allows 1:1 Aspect Ratio. Despite this, the results are often impressive and hold great potential for future developments.
Thanks for reading!
Hi👋 I’m Erik. If you liked this story, clap and follow me here on Medium.
I‘m sharing my learnings in my newsletter.
It would be amazing if you could join me on this journey.