- Character.ai’s new Avatarfx tool can make a single photo for a realistic video
- The video shows a still image that speaks with synchronized lip movements, movements and long -shaped performances
- The tool supports everything from human portraits to mythical beings and speaks lifeless objects
Pictures may be worth a thousand words but character.ai see no reason why the image should not speak these words themselves. The company has a new tool called Avatarfx that transforms still images into expressive, speech, singing, gesticulating video avatars. And not just photos of humans, animals, paintings of mythical animals, even a lifeless object can speak and express feelings when you include a voice test and manuscript.
Avatarfx produces surprisingly compelling videos. Everything from lip synchronization accuracy, nuanced main slopes, eyebrows raise and even appropriate dramatic hand movements is everything there. In a world that is already swirling with AI-generated text, images, songs and now whole podcasts, Avatarfx may sound like just another smart toy. But what makes it special is how smooth it connects to voice to visuals. You can feed it with a portrait, a dialogue line and a tone and character.ai calls what comes out of a performance, one capable of long -shaped videos too, not just a few seconds.
It is thanks to the model’s temporal consistency, a fancy way of saying that the avatar suddenly does not grow a third eyebrow between sentences or forget where its chin goes in the middle of the monologue. The movement of the face, hands and body synchronization with what is being said, and the final result looks, if not alive, then at least lively enough to play in a late-night infomercial or guest host a podcast about the room goals. Your creativity is the only thing that stands between you and an AI-generated soap opera starring the family refrigerator. You can see some examples in the demo below.
Avatar alive
Of course, the magic talking picture frame fantasy comes with its reasonable proportion of luggage. An AI tool that can generate lifelike videos raising some understandable concerns. Character.ai seems to take these concerns seriously with a package of built -in security measures for Avatarfx.
It includes a ban on generating content from images of minors or public persons. The tool also crawls human uploaded faces so that they are no longer accurate and all scripts are controlled too appropriate. Should that not be enough, each video has a watermark to make it clear that this is not real recordings, just some impressively animated pixels. There is also a strict policy with a strike to break the rules.
Avatarfx is not without a precedent. Tools such as Heygen, Synthesia and Runway have also pushed the boundaries of AI Videor Reneer. But Character.ai’s entry into space UPS Ante by melting expressive avatars with its signature chat personalities. These not only speak heads; They are characters with back stories, personalities and the ability to remember what you said the last time you talked.
Avatarfx is currently in a test phase where character.ai+ subscribers probably get first DIBS when it rolls out. For the time being, you can join a waiting list and start dreaming about which of your friends’ selfies would make the best Shakespearic monologue delivery system. Or what version of your childhood -outdated animal that can finally become your therapist.
