Character.AI has introduced a new AI model called AvatarFX, allowing users to create videos where the avatar can speak, sing, and express emotions based on input images. The results can range from realistic human portraits to cartoon characters in animation.
Creating videos with moving mouth or motion in images may not be new, but Character.AI highlights the strengths of AvatarFX in its ability to support various inputs, including 2D, 3D images, and even animal faces. The facial features remain consistent and support the creation of long videos.
In terms of security, AvatarFX embeds watermarks to indicate that the video is not real footage. It also includes safeguards to prevent the upload of minors, politicians, or public figures’ images. Personal images are edited to ensure some details are not identical, and users must agree not to misuse AvatarFX for illegal or inappropriate purposes.
Character.AI+ customers will have early access to some of these capabilities.
(Source: Character.AI)
TLDR: Character.AI introduces AvatarFX, an AI model enabling users to create videos where avatars can speak, sing, and display emotions using various input images, from human portraits to cartoon characters. The tool includes security features to prevent misuse and provide access to certain users.
Leave a Comment