AIVIA — One capture. Infinite direction.

How it works

Three steps between a bare studio capture and a finished campaign. The human presence stays real; everything around it becomes directable.

Capture

STEP 1

Shoot once, neutral

A real model, photographed flat and even on a green or blue seamless screen — wearing the real garment, carrying the real product. Bare face, natural hair, unretouched skin. The less styling in the plate, the more the model can direct later.

Train

STEP 2

Train the model

Our own AI model learns the captured identity and your brand's visual language — product, palette, mood. It is trained on your material, not scraped stock.

Generate

STEP 3

Generate the campaign

Face styling, hair, environment and cinematic lighting are composed to your brief. The garment stays exactly as captured — new set, new market, new format without another day on set.

Our trained model

Generic image generators invent people who don't exist and drape fabric the way no real garment ever falls. AIVIA works differently: a controlled capture in, a controlled campaign out.

Face & hairStyled, groomed and lit to brief — while the underlying identity remains consistent frame to frame.
Never the productGarments, bags and accessories are never generated. Fabric only falls naturally on a real body — so the model wears the real piece, and it stays untouched.
Background & setConcrete minimalism, sunlit interiors, location looks — composed, not booked.
Lighting & gradeCinematic key, editorial softness, campaign-grade color — relit from the neutral plate.

Trained model Green / blue screen Identity-consistent

One plate, four campaigns

The proof is repeatability. From a single capture, the same identity and the same real garment carry across set and lighting changes — the control a brand needs to run imagery at scale.

Real model in a crimson silk dress on a blue chroma-key screen in a photo studio — Input plate · chroma blue

Same model and dress on a rooftop at violet dusk — Editorial · violet hour

Same model and dress in a golden sunlit luxury interior — Campaign · warm light

Same model and dress in a dark studio under a single spotlight — Lookbook · studio noir

Same model and dress on a bright mediterranean coastal terrace — Resort · coastal day

Same model, same real garment in every frame. Different environment and light in each — generated, not reshot.

Built for volume

E-commerce refreshes, seasonal campaigns, market-specific creative — once the model is trained, output scales without another day on set.

Campaign frame: ivory gown in a brutalist concrete gallery

Campaign frame: crimson dress on a neon-lit night street

Campaign frame: tailored black suit in a white studio

Campaign frame: crimson dress on desert dunes at sunset

Campaign frame: ivory gown under a spotlight in a dark gallery

FormatsCampaign, e-com, social crops from one master frame

TurnaroundFinished frames in hours from an approved brief

RightsContracted capture, clear usage — no stock ambiguity

Request a demo

See your brand
through the model.

Tell us who you are and what you sell. We'll walk you through the capture process, show live input→output pairs, and scope what a trained model for your brand would deliver.

Prefer email? Write to dr@deras.de