APT AI Video Generator: Turn Two Photos into a ROSÉ & Bruno Mars-Inspired Dance

Make an APT-inspired music video in AnyMimic by assigning two photos to the performers in your reference. Reimagine the playful ROSÉ and Bruno Mars duet with your own cast, using a pink-stage clip to guide the scene and motion.

0/1
0/1
0/1
선택 Ref video· 또는 내 파일 업로드
선택 사항 - 0/1

0/4000

How to Create an APT Duo AI Video in AnyMimic

Treat casting and choreography as separate inputs: your images choose the people, while the APT video provides their performance. Choose the APT example to apply role instructions tailored to its female and male performers.

1

Choose a Clear Two-Person MV Segment

Choose APT from the video examples, then select a segment of its 22-second source in the trimming dialog. Paid reference input is limited to 15 seconds and free input to 3 seconds. You can also try Hotel Lobby or supply a suitable clip of your own.

2

Assign the First and Second Character Images

With the APT example selected, place the replacement for the female performer in Character Image 1 and the replacement for the male performer in Character Image 2. This assignment follows the original person, even when the camera changes their screen position.

3

Confirm the Identity Mapping across the Sequence

Check the custom APT instructions for the female-to-image-1 and male-to-image-2 mapping across the sequence. Confirm that the selected segment shows the interaction you want, then choose output settings and review the displayed credit estimate.

4

Render and Check the Duo Motion

Generate the clip, then watch the turns, close-ups, and shared gestures for identity drift. Download a usable take and finish any music or captions in the app where you plan to share it.

Why Two-Person MV References Need Careful Casting

A convincing AI APT Video depends on more than recognizable faces. The cast needs to remain understandable as the camera and performers move through the same scene.

Screen Position Is Not Permanent Identity
A performer who starts on the left can appear on the right after a cut. Mapping photos only by position can become ambiguous when a routine changes the composition.
Tight Portraits Leave Out Appearance Details
An image with a tiny or hidden face gives the model less to work with. Visible facial features, hair, and clothing make it easier to evaluate a replacement role.
Different Backgrounds Can Compete
Portrait locations and the performance set may have different lighting or surroundings. The replacement instructions use the video environment while taking character appearance from the images.
Image 1

Develop an AI APT MV with a Cast You Can Review

Start with a clear duo concept and evaluate a short take before extending your edit. This workflow is useful when you want to test how two people or original characters fit a shared performance.

Control Who Takes Each Role
Use the two upload slots to make casting deliberate, whether you are creating a friends duet, a couples clip, or an original character pairing.
Change One Input at a Time
Keep your reference fixed while trying a different photo for one performer. Comparing short drafts can help you identify which input needs improvement.
Develop a Recurring Duo
Reuse character images you own across suitable performance references. Check each output for consistent appearance rather than assuming every generation will match.
Inspect Difficult Moments Early
Use a short segment to evaluate face turns, overlapping bodies, and quick exchanges before committing more credits to a longer result.
Match Your Intended Framing
Choose portrait, square, or landscape from the supported settings. Look at both characters in the preview to check that the selected frame serves the duet.
Plan the Soundtrack Separately
Decide whether to request generated audio or finish the music after downloading. This keeps the visual performance and your publishing soundtrack choices clear.

APT AI Video Generator Questions

Practical answers about performer assignment, references, settings, and finishing your clip.