1
Upload Portrait Image & Subject Detection
Choose an image with a person; after upload the system runs subject detection to verify eligibility.
No image uploaded yet.
Before subject detection passes, the Generate Video button remains disabled.
2
Voice Selection (Clone / Use Existing Voice)
Choose one: upload audio to train a new voiceprint, or enter an existing speaker_id.
No training audio uploaded yet.
Current speaker_id:
(none)
3
Generate Script From Knowledge Base & Synthesize Audio (use Step 2 voice)
Upload one or more documents (txt/pdf/docx/doc) to extract text as a knowledge base. The model generates a script (editable). Click Synthesize Audio to use it for video generation.
No script generated yet.
4
Generate Digital Human Video
After subject detection passes and audio is ready, click to submit a video generation task and poll its status.
Waiting for subject detection to pass...