Flat-Lay & Ghost Mannequin to Runway Video: The 2026 AI Workflow

Short answer: Start from a clean, well-lit flat-lay or ghost-mannequin photo. First turn it into one approved on-model still, then animate that still into a short video. Going image → approved still → video keeps the garment and the face far more consistent than asking a video model to do everything at once.

The three-step pipeline

Step	Input	Output	What to check
1. Clean the source	Flat-lay or ghost mannequin photo	Even light, true colour, no wrinkles	Colour matches the real sample
2. Put it on a model	Source photo + model brief	One approved on-model still	Seams, print scale, logo, fit
3. Animate	Approved still	5–10 second walk, turn or campaign clip	Fabric movement, face drift, hands

Step 1 — Prepare the garment photo

• Shoot flat on a neutral background with soft, even light.
• Photograph print and logo details separately as close-ups for reference.
• Keep the colour true: AI will copy a colour cast, not fix it.

Step 2 — Create the on-model still

Prompt formula:

[Garment description, fabric and colour] worn by [model brief], [pose], [setting], [lighting], keep the exact print, seam placement and logo from the reference, editorial e-commerce photography.

Generate several, choose one, and treat it as your approved frame. Every later step inherits from it.

Step 3 — Animate the approved still

Pick one simple movement: walk toward camera, slow turn, or a slow push-in. Describe how the fabric behaves — a linen shirt moves differently from a wool coat.

The model walks slowly toward the camera, the [fabric] moves naturally with each step, camera slowly pushes in, studio light stays constant.

Common problems and fixes

• The print changes → shorten the clip and reduce camera motion.
• The face drifts → always start the video from the approved still, never from text alone.
• Fabric looks stiff or plastic → name the fabric weight and drape in the prompt.

Where to do this

• Garment to Campaign in Motion Studio — upload one garment photo and pick a scene.
• AI Ambassador Studio — keep the same model face across products.
• Pay-as-you-go studios — no subscription, top up from €5.

Learn the consistency method step by step in the AI Consistency: Character & Product workshop.

FAQ

Can AI turn a flat lay into a fashion video? Yes. The most reliable way is flat lay → approved on-model still → short video from that still.

Does a ghost mannequin photo work better than a flat lay? Often yes, because it already shows the garment's 3D shape, which helps fit and drape.

Will the result be accurate enough for e-commerce? It can be, but a person should always compare it with the real sample before publishing.