Jason Miles
The well-known social media network TikTok recently declared that it would be integrating generative AI avatars. Users will be able to design and develop their own virtual personalities with these artificial intelligence-powered avatars.
Computer-generated characters known as “generative AI avatars” can be created to resemble any object the user wants. They are produced by an AI that uses a combination of algorithms and deep learning methods to produce intricate, one-of-a-kind designs depending on user preferences.
Users will have a variety of personalization options with TikTok’s new generative AI avatars, including choices for skin tone, hairstyle, attire, and accessories. The animated avatars will be capable of a variety of expressions and behaviours, including lip-syncing, dancing, and singing.
TikTok has collaborated with some of the top AI and computer graphics specialists in the world, including the University of California at Berkeley, and Carnegie Mellon University, to develop these avatars. NVIDIA created a neural network dubbed StyleGAN, which serves as the foundation for the avatar technology.
A generative adversarial network (GAN) is a sort of network that comprises of two neural networks that are in competition with one another. While the other network tries to differentiate between authentic and fake images, the first network creates images. The networks learn from one another over time, increasing the realism of the images that are produced.
Although TikTok is still developing its generative AI avatars, it has already made a demo video showcasing some of the features. The user is seen in the video creating an avatar and then engaging in a variety of activities, including dancing and playing musical instruments.
For the platform and its users, the addition of generative AI avatars to TikTok is a promising move. It opens up a completely new world of potential for creativity and self-expression and may pave the way for even more sophisticated AI-powered features in the future.
