Step inside the story.
An educator chooses a story. Their digital character transforms scene by scene while the background responds to each setting. The camera captures real movement, so children see the character teach inside the story world.
🏹 Scene 1 · Munich city square sunny sun
Hallo! Willkommen in München!
Hello! Welcome to Munich!
Educator · motion capture
Markerless360° character preview Text → 3D
Drag to rotate · separate previewThis character was written into being
The witch educator on the left began with one prompt: “a cute cartoon witch teacher for children, full body, friendly smile”. AI generated a complete 3D model and it renders live in your browser. An educator can imagine any character and create it with a sentence.
Production pipeline
- ① Scene description → real-time background (✓ live)
- ② Character setting → twin transforms scene by scene (✓ live)
- ③ Camera pose stream drives character (✓ live)
- ④ Prompt → 3D character generation (✓ live)
- ⑤ Main stage: real educator movement drives 3D bones (✓ live)
This area previews the model’s appearance. The real 3D twin on the story stage above is what follows the educator’s movement.
Immersive
Backgrounds move with the story and characters transform with the plot — children are not watching a lesson; they are living in a story.
The educator becomes the performer
An ordinary camera replaces wearable motion-capture equipment, turning each teaching gesture into live character animation.
A path to generative worlds
With image and 3D generation APIs, scenes and characters can grow into cinematic worlds. The pipeline is ready.
Privacy note: video and pose data are processed locally in your browser and are never uploaded. The AI engine (MediaPipe Pose) is hosted by MoliVerse. Current scenes and characters are generated procedurally in real time; richer AI image backgrounds and text-to-3D characters can be connected through partner APIs in a future phase.