Discover how Presti uses Generative AI to revolutionize furniture industry photo shoots, saving costs and enhancing product presentation for new collections
The process is taxing and costly due to the difficulty of relocating furniture.
That is why Presti, a French startup established in November 2022, employs generative AI to transform a single product image into a genuine lifestyle photo. The company recently secured a $3.5 million seed round led by the global tech investment firm Partech. Additionally, numerous business investors have contributed to the funding.
Nabil Toumi, the CEO and co-founder of Presti AI, stated to TechCrunch, “We promptly contacted 50 potential users through the phone.” “And everyone said the same thing.” Producing product visuals was complicated, time-consuming, and expensive, and there was no straightforward solution for these photographs. Simultaneously, it was the most critical brand asset in establishing a distinctive identity and selling online.
Presti transforms this:
Into this:
Initially, Presti did not restrict its scope to furniture manufacturers. However, the startup promptly identified that furniture manufacturers encountered several notably distressing challenges.
“They had high logistics costs associated with a photoshoot, as they needed to rent a nice house and transport products.” “And those photo shoots were organized months in advance, resulting in expenses of hundreds of thousands, if not millions, of euros annually,” Toumi stated.
Presti’s fundamental model is Stable Diffusion XL. It has been retrained and optimized to perform exceptionally well in the furniture industry’s product imagery.
Initially, the team attempted to employ a standard version of Stable Diffusion XL. However, they promptly identified the problems. “The backrest will be distorted, and legs will be affixed to your sofa,” Toumi stated. In the same vein, it was challenging to achieve the appropriate perspective. For example, the wall behind the sofa must be perpendicular to the sofa.
“At the same time, the dataset on which we trained our model is paramount.” Toumi stated, “We have over 75,000 ultra-high-quality furniture images in our industry that we can use to train our model and reinforce the learning process for a specific use case, specifically for this type of photo.”
Presti was unwilling to conclude at the backdrop generation stage. Additionally, customers have the option to include accessories. For example, cushions may be implemented when creating product images for a new sofa. These cushions will cast a realistic shadow on the settee, ensuring they do not appear to have been inserted in Photoshop.
In the same vein, furniture manufacturers frequently offer numerous variations of the same model, each with its unique texture and color. Although this is still a work in progress, Presti anticipates that its consumers can exchange the material using its tool. Consequently, introducing new products will be significantly simplified for companies that utilize Presti.
Conversely, freelance photographers are unlikely to appreciate this new product. Whether the creativity and originality of other skilled humans who may work on a physical photo shoot, such as location stylists, can be wholly replaced through machine-generated backgrounds without the resulting artificial lifestyle images looking a bit samey is one open question.
Although Presti primarily collaborates with mid-sized furniture manufacturers, it also maintains a strategic partnership with Maisons du Monde, one of France’s largest furniture retailers. Maxime Brousse, Thibaud Elzière, Julien Hirth, Abou Laraki, and Rémi Lemonnier are among the other investors in the venture, in addition to Partech.