Preparing a retailer’s online catalog once required expensive physical photo shoots to capture products from all angles. A Tel Aviv startup saves brands time and money by turning those camera clicks into mouse clicks.

Hexa uses GPU-accelerated computing to help businesses turn their online inventory into 3D visualizations that shoppers can view in 360 degrees, animate or even experience virtually to help them make a buying decision. A company that recently announced a $20.5 million funding roundcooperates with brands of fashion, furniture, consumer electronics, etc.

“The world is becoming 3D,” said Yehiel Atias, CEO of Hexa. “Just a few years ago, the digital infrastructure for this was so expensive that it was more affordable to arrange a photographer, models and lighting. But thanks to advances in artificial intelligence and NVIDIA GPUs, retailers can now use synthetic data to replace physical photo shoots.”

Hexa’s 3D renderings are used by major retail websites such as Amazon, Crate & Barrel, and Macy’s. The company creates thousands of renderings every month, reducing the need for physical photo shoots of every product in a seller’s catalog. Hexa estimates that this can save customers up to 300 pounds of carbon emissions for each product imaged digitally instead of physically.

From physical photo sessions to visualizations with artificial intelligence

Hexa can reconstruct a single 2D image or a set of low-quality 2D images into a high-fidelity 3D resource. The company uses different levels of automation for its renders, depending on the complexity of the shape, the amount of visual data that needs to be reconstructed, and the similarity of the object to the existing Hexa dataset.

To automate elements of their workflow, the team uses dozens of artificial intelligence algorithms that were developed using the PyTorch deep learning environment and run on NVIDIA Tensor Core GPUs in the cloud. For example, if one of Hexa’s artists is reconstructing a 3D toaster, one algorithm can identify similar geometries the team has created in the past to give the creator an edge.

Another neural network can scan a retailer’s website to determine how many of its Hexa products support 3D visualization. The company’s entire pipeline also runs on NVIDIA GPUs available through Amazon Web Services.

“Access to computing resources through AWS gives us the ability to use thousands of NVIDIA GPUs at any given time,” said Segev Nahari, Hexa’s lead technical artist. “If I need 10,000 frames ready by a certain time, I can request the hardware I need to meet the deadline.”

According to Nahari, rendering on NVIDIA GPUs is 3x faster than CPUs.

Going beyond retail, venturing into the universe

Hexa developers are constantly experimenting with new 3D rendering methods — looking for workflow improvements in preprocessing, object reconstruction, and postprocessing. The team recently started working with NVIDIA GET3Dgenerative AI model from NVIDIA research which generates highly accurate 3D shapes based on a training dataset of 2D images.

sneakers created by GET3D
By training GET3D on the Hexa shoe dataset, the team was able to create 3D models of new shoes that are not part of the training data.

In addition to its e-commerce work, Hexa’s research team explores new applications for the company’s AI software.

“It’s not limited to retail,” Atias said. “Industries from gaming to fashion and healthcare are finding that synthetic data and 3D technologies are a more efficient way to do things like digitize inventory, create digital doppelgangers and train robots.”

The team counts its membership in The beginning of NVIDIAa global program that supports cutting-edge startups, as a “huge advantage” in improving the technology Hexa uses.

“Being a part of Inception opens doors that outsiders don’t,” said Atias. “For a small company trying to navigate NVIDIA’s huge range of hardware and software offerings, this opens the door to all the cool tools we wanted to experiment with and understand the potential they could bring to Hexa.”

Hexa is testing NVIDIA Omniverse Enterprise platform—an end-to-end platform for creating and managing metauniverse applications—as a tool to unify the annotation and rendering workflows used by dozens of 3D artists around the world. Omniverse Enterprise enables geographically distributed creative teams to customize their rendering pipelines and collaborate to create 3D assets.

“Each of our 3D artists has a different software process they’re used to, so it can be difficult to get a unified output while remaining flexible with the tools each artist uses,” said Jonathan Clarke, CTO of Hexa. “Omniverse is an ideal candidate in this regard, with huge potential for Hexa. The platform will allow our artists to use whatever rendering software they prefer, while also allowing our team to render the final product in one place.”

To learn more about NVIDIA Omniverse and new generation content creationsign up for free for NVIDIA GTCa global conference on the era of artificial intelligence and the metaverse, which will take place online March 20-23.

Image and video courtesy of Hexa