Vestiaire Collective standardizes luxury resale photos with AI

The company turned to Photoroom after finding that most generative AI tools altered product details, such as stitching or logos, which complicated their digital authentication process. Because Vestiaire Collective verifies items based on photos, preserving the fidelity of the original item is a requirement for maintaining buyer trust. Photoroom’s system, which leverages open models like FLUX.1 and Meta’s DINOv2, ensures that shadows and lighting remain consistent without distorting the physical reality of the goods.

Thomas Hezard, Chief Product Officer at Vestiaire Collective, noted that the platform previously relied on an in-house clipping tool because existing market solutions lacked the necessary precision. The new implementation, which took eight weeks to deploy, now processes images via an API that automatically centers items and adjusts shadows based on the product category. This technical upgrade allows the platform to maintain high-speed operations while ensuring that product pages remain accurate enough for both authentication and marketing efficiency.