Fashion: how AI is transforming flat catalog images into dynamic assets

Fashion brands turn to AI to transform flat photos into on-model images. A crucial shift for resale, where missing archives and the need for buyer reassurance remain major hurdles.

‍

Image, the currency of e-commerce

In fashion e-commerce, the image is often worth as much as the product itself. Yet in resale, one persistent problem undermines the experience: past catalog visuals rarely survive. When a garment comes back on the market, what’s left is usually a flat shot or a hanger photo. These unengaging images deprive shoppers of vital cues—fit, drape, fabric. The result is hesitation at checkout, just as resale is entering a phase of unprecedented growth.

The global resale market is projected to reach $200 billion by 2030 (ThredUp), growing twice as fast as new retail. But profitability hinges on how effectively brands can elevate the visual value of their products.

‍

From packshot to on-model

AI-powered “on-model generation” solutions promise to solve this challenge. From a flat or hanger image, the algorithm rebuilds the garment and projects it onto a virtual model.

-The process involves three stepSegmenting the item and correcting imperfections.

-Reconstructing the shape and adapting it to a body.

-Rendering the final image, including model selection (skin tone, size, style) and visual harmonization.

For brands, the benefits are twofold: streamlining heterogeneous archives and reducing dependence on expensive photo shoots. Some platforms claim up to 75% lower production costs and a fivefold faster time-to-market.

‍

A fast-moving ecosystem

Start-ups are racing to capture this space:

-Botika (Israel) turns flat shots directly into on-model visuals with a library of virtual models.

-Vue.ai (US) targets retailers with complete automation workflows.

-ZMO.ai (China) and Huhu.ai focus on high-volume processing through APIs.

-Veesual (France) emphasizes customer experience by letting shoppers switch models directly on product pages.

-Lalaland.ai (Netherlands) partners with Levi’s to enhance model diversity.

Major retailers are scaling adoption. Zalando reports that nearly 70% of its recent visuals were generated with AI. H&M has experimented with virtual models in campaigns, while Levi’s has tested AI-generated avatars to boost inclusivity.

‍

A strategic lever for resale

Beyond productivity gains, the impact on buying intent is measurable. Research cited by Veesual suggests the likelihood of adding to cart doubles when consumers see on-model visuals. In resale, where reassurance is critical, the effect is even stronger.

AI also addresses a structural issue: the disappearance of iconographic archives. By “reanimating” past collections, it gives new commercial value to thousands of items that would otherwise remain visually unappealing.

‍

Limits and open questions

The technology raises three concerns:

-Accuracy: Any misrepresented color or fit risks disappointing customers and increasing returns.

-Artificial diversity: Virtual inclusivity may be dismissed as cosmetic if real-life campaigns fail to reflect the same values.

-Transparency and regulation: The EU is considering rules on labeling AI-generated content, leaving brands to decide whether to disclose it explicitly.

‍

Toward a new standard for circular fashion?

As resale reshapes the industry, brands are rethinking both operations and tools. AI-driven image generation could become, within years, as standard as e-commerce photo shoots became in the 2000s.

By bridging lost archives and today’s visual expectations, these technologies deliver a pragmatic answer to a fast-changing market: selling faster, at lower cost, while giving buyers the confidence they need.

‍

‍

‍

‍