AI face replacement for OTT microdrama
Upload a photo and step into an existing drama scene. The actor gets your face. The suit, the set, the subtitles and the other actors stay.
Your photoReal consenting person, unedited selfie. Final output at 720p. Source audio only.
How it works for the user
One selfie is enough. The service cuts it to a clean face oval and removes hair, neck, clothes and background.
Choose an existing microdrama scene and the actor you want to replace.
The scene comes back with your face in the role. Reviews run before the download.
Best results
Press play. Each group plays in sync, with the original audio.
Target: the bald man in the light suit. His face changes. The suit, menu, subtitles and the woman's one-frame shot stay in the final 720p output.
Same clip, same selfie, same prompt. The early 480p draft also replaced the face, but lost the woman's frame.
A public stock photo goes through the full service. The face and beard visibly change. The user considers this replacement correct.
The photoThe first real result, from the P1 prototype before the current service. Expressions are mostly correct, with mild exaggeration.
The photoWhy it matters for an OTT platform
The service works on video you already own. These are the possibilities it opens. They are not measured results.
Any catalog scene could get a version where the viewer plays the lead.
Seeing yourself in the story could bring viewers back to the next episode.
A short scene with your own face is content that viewers may want to send to friends.
The input is existing footage and one photo. The set, cast and subtitles stay as they are.
For developers
One REST API covers the whole flow: upload a photo, pick a scene, create a job, follow progress and download the result. Bearer-token authentication and JSON throughout.