Technology
Presti uses GenAI to replace expensive photo shoots in the furniture industry
If you’ve ever bought a settee online, have you ever ever considered the houses you see in the background of product photos? When it comes time to launch a brand new collection, furniture brands typically spend a small fortune on photo shoots. It’s a tedious and expensive process, since it’s hard to move furniture around.
That’s why a French startup called PriestFounded in November 2022, uses generative AI to transform a single product image into a practical lifestyle photo. The company just raised $3.5 million in a seed round led by global investment firm Partech, which also includes several angel investors.
“We picked up the phone really quickly and talked to 50 potential users,” Presti co-founder and CEO Nabil Toumi told TechCrunch. “And they all said the same thing. Creating product visuals was a very long process, it cost a lot of money, and they didn’t have a simple solution for creating those images. At the same time, it was really the most important asset for brands to create a unique identity and sell online.”
Presti puts it this manner:
Involved in this:
Initially, Presti didn’t limit its focus to furniture corporations. However, the startup quickly realized that furniture corporations face some particularly difficult problems.
“They had to rent a nice house for the photo shoot, they had to transport the products — so they had high logistics costs. And these photo shoots were planned months in advance and ended up costing them hundreds of thousands, even millions of euros a year,” Toumi said.
At its core, Presti uses Stable Diffusion XL as its base model. It has been trained and modified to perform particularly well for product images in the furniture industry.
At first, the team tried using the regular Stable Diffusion XL. But they quickly realized there have been some issues. “The sofa would have legs added to it, and the back would be distorted,” Toumi said. Similarly, getting the perspective right was difficult. For example, the wall behind the sofa had to be parallel to the sofa.
“At the same time, something that’s really important is the dataset that we trained our model on. We currently have over 75,000 images of ultra-high-quality furniture images in our industry that we can use to train our model to strengthen the learning process for the specific use case, for these types of images,” Toumi said.
Presti didn’t want to stop at generating backgrounds. Customers also can add accessories. For example, if you happen to’re generating product photos for a brand new sofa, you may add pillows. These pillows will forged realistic shadows on the sofa, so that they won’t seem like something added in Photoshop.
Similarly, furniture brands typically have several variations of the same model with different textures and colours. While it’s still a piece in progress, Presti hopes its customers will give you the chance to swap materials using its tool. As a result, it should be much easier for corporations using Presti to release latest products.
On the other hand, freelance photographers won’t like the latest product. And whether the creativity and originality of other expert individuals who might work on a physical shoot, akin to on-location stylists, will be completely replaced by machine-generated backdrops—without the resulting artificial lifestyle images looking a bit the same—is an open query.
Although Presti works mainly with mid-sized furniture corporations, it also has a strategic partnership with Maisons du Monde, one in all the largest furniture retailers in France. In addition to Partech, other investors in the startup include Maxime Brousse, Thibaud Elzière, Julien Hirth, Abou Laraki and Rémi Lemonnier.