Vertical video gen
Vertical video generation is where the locked character starts moving. This lesson shows you how to download a reference video, prepare the influencer image, and combine both inputs in a motion transfer tool.
You will leave withOne approved reference video, one prepared model image, and one vertical motion-transfer clip that still reads as the same character.
Turn the influencer image into video
Video generation starts with two inputs: a reference video for motion and the influencer image for identity.
Reference video
Find a social media clip with the pose, gesture, pacing, and framing you want to transfer.
- Best for
- Supplying motion and camera energy.
- Inputs
- Downloaded vertical clip
- Output
- A clean motion reference.
Influencer image
Use the approved still image from Character creation or Character modification as the person to animate.
- Best for
- Preserving the generated influencer identity.
- Inputs
- Approved still image
- Output
- The model image for video.
Motion transfer tool
Upload both inputs into the motion transfer option and let the tool create the influencer video.
- Best for
- Rebuilding the motion with your own generated person.
- Inputs
- Reference video + Influencer image
- Output
- A vertical influencer clip.
Choose and download the reference video
Start from a clean social media video. The better the reference movement is, the easier the final clip is to judge.
Reference video
Video: choose the reference video
Show how to pick a simple vertical clip, download it, and check that the motion is readable before generation.
Motion transfer
Video: run motion transfer
Show the reference video and influencer image being uploaded into the motion transfer option to create the final video.
Motion transfer run brief
Use the downloaded reference video only for motion, timing, gesture, camera energy, and framing. Inputs: - Reference video: [describe the movement, pacing, and 9:16 framing]. - Influencer image: [approved still image from Character modification]. Motion goal: - Transfer the reference movement onto the influencer image while keeping the influencer's identity stable. Keep stable: - Face identity, hair silhouette, body proportions, wardrobe logic, and vertical composition. Avoid: - Face drift, limb warping, background collapse, over-fast motion, identity loss, and unnatural camera shake.
Task: download one clean reference video, prepare the influencer image, then run the motion transfer tool with both inputs.
Reference video + influencer image
The reference video provides the movement. The influencer image provides the person. The output should feel like that same person performing the reference motion.
Reference video
A downloaded vertical clip from social media with a clear action and usable framing.
Our influencer image
The approved still image from the earlier character workflow, ideally matching the video style.
Motion transfer
Use the motion transfer option to combine the reference video with the influencer image.
Influencer video
A vertical video where the generated influencer follows the reference motion without losing identity.
Review the generated video
Do not approve the clip only because it moves. Approve it when identity, body shape, timing, and framing all survive the transfer.
Reference video
Supplies the action, pacing, and vertical framing.
Influencer image
Supplies the face, styling, and character continuity.
Influencer video
The final clip should pass both the motion test and the same-person test.
Vertical motion transfer run brief
Use this before launching the clip. It keeps the reference, source image, and motion goal aligned before you run the motion transfer tool.
Use the downloaded reference video only for motion, timing, gesture, camera energy, and framing. Inputs: - Reference video: [describe the movement, pacing, and 9:16 framing]. - Influencer image: [approved still image from Character modification]. Motion goal: - Transfer the reference movement onto the influencer image while keeping the influencer's identity stable. Keep stable: - Face identity, hair silhouette, body proportions, wardrobe logic, and vertical composition. Avoid: - Face drift, limb warping, background collapse, over-fast motion, identity loss, and unnatural camera shake.
What you should finish with
This topic is complete when these outputs exist and are saved for the next stage of the workflow.
- One approved reference video for the motion transfer run.
- One prepared model image created through the Character modification workflow.
- One approved vertical clip created with the motion transfer tool.
- One short review note explaining whether the clip preserves identity well enough for further scenes.
Placeholders for uploads
These are the assets we will plug in later. Keeping the slots visible now makes the workflow feel complete and shows exactly what still needs to be collected.
placeholder
Vertical reference video
Upload the reference clip that supplies the motion pattern for the final vertical generation.
placeholder
Prepared model image
Upload the approved modified still image used as the source for motion transfer.
placeholder
Motion transfer output
Upload the generated vertical clip that came from the motion transfer run.
placeholder
Vertical clip review note
Upload the short review sheet used to judge identity stability, framing, and motion quality.
Once one vertical clip is stable, Transition video gen is where you start chaining multiple scenes together into a more finished piece.
Continue to Transition video gen