What “two-photo duo” means
A two-photo duo workflow gives the model one image for each intended character. In this generator, Human Duo is fixed to two images and a vertical 9:16 output. The prompt asks for synchronized movement, consistent identity, and neon stage lighting. It does not accept an arbitrary text prompt or a user-selected reference video.
This setup is useful when you want two consenting adults in the same short scene. It is not a tool for creating a celebrity, impersonating a public figure, or recreating an A2O LITE member. The page and server both require the adult-permission confirmation.
Choose clear, compatible images
Upload one photo per person, with faces and enough of each person’s body visible for the planned dance movement. Even lighting, a simple background, and a similar camera distance can make the pair easier to combine. The interface accepts JPG, PNG, WebP, and HEIC files up to 10 MB each.
Preview both selections before generating. Remove or replace a file if a face is hidden, the image belongs to someone else, or the person did not agree to use it. Do not include children, public figures, or recognizable A2O LITE members. The system may block images that appear uncertain or do not fit the selected template.
Confirm permission before uploading
The Human Duo consent statement asks you to confirm that each person is 18 or older, has given permission, and is not a public figure or A2O LITE member. This is a product restriction, not a guarantee that every image can be automatically classified correctly. If an image cannot be safely reviewed, the system rejects it.
Keep permission records outside the photo itself and never upload identity documents. After sign-in, the selected images are sent to the configured storage route and the generation service. They are used to produce the requested video; read the privacy policy for storage, provider processing, and account deletion information.
Start and follow the generation task
The request contains the template name, image URLs, fixed aspect ratio, 15-second duration, and selected resolution. The server replaces any client attempt to alter the prompt or reference video with the approved template values. The task is placed in the existing queue and linked to your account.
The page polls the task every five seconds. You can see when it is queued, processing, and ready. The provider may take several minutes, and a progress bar is only an estimate. If you close the page, open My videos after signing back in to check the task and download its MP4.
Understand resolution and credits
The form offers 720p and 1080p. The server calculates the credit requirement from the selected template, duration, and resolution rather than trusting a number sent by the browser. A higher resolution uses more credits under the current product settings. The product screen displays the estimate before submission.
The one-time packs contain credits for the listed number of standard video generations. If a task fails or is cancelled, the task service restores the associated generation charge. Check My videos and your balance before repeating a slow request.
Output and sharing
A successful job returns an MP4 preview and a download action. The completed task is associated with your account and can be opened again from My videos. Review the result for visual artifacts, accidental likeness, or unwanted background details before you share it.
Describe the clip as AI-generated fan content. It is not official A2O LITE choreography or a real group performance. The website does not attach the “NEON” recording, provide lyrics, or grant permission to use music in a post.
HowTo: create a two-photo dance
Follow the same steps in the generator above.