ToThree.io
Turn a selfie into a 3D avatar.
Mobile Application
Serves as a gateway for new users. Once they create an avatar here, this is the hub where they can manage their virtual life. Of course, the same app is available on Desktop. Probably Flutter.
Take a selfie, and we will generate it into an avatar.
We might even take 1-3 business days to generate the avatar, for that "exclusive" experience. This would save us processing.
Avatar Filter
Users can have their avatar act as a filter for videos. When they are recording themselves, such as for Youtube videos, their avatar can be speaking instead of them.
Because we make 3D avatars, we can also have the avatar spin & move around in the space.
We might add the ability to design the avatar's space, although there are complications here regarding the person's video.
How do we identify the face, and place it accordingly in the virtual room?
Do we simply replicate exactly what the actual video is showing, but with the avatar?
Aritificial Intelligence
When the person takes the selfie, we use Stable Diffusion to gloss up the facial features.
We remove background noise and make the person more beautiful.
Perhaps, we even have the AI specifically position the face in the picture, and guess the hidden facial features.
We will need to see or guess the back of their head.
For that, we might need to have the user submit multiple photos, or a 360 view of their head.
Or, perhaps, full body. It all depends on the capabilities of the AI model we use.
Then, we turn that image into a 3D avatar. There must be pre-exsiting AI models for this process.
Last updated