AI in Retail: Israeli Startup Hexa is Enabling Shoppers to View Products in 360 Degrees & Try Them Virtually
3D asset visualization and management platform, Hexa is transforming the way retail brands and businesses are showcasing and selling their products. The Israeli startup employs GPU-accelerated computation to assist businesses in creating 3D representations of their online goods that customers can examine in 360 degrees, animate, or even try on digitally.
Recently, the Tel-Aviv-based startup raised a $20.5 million series A round led by Point 72 ventures and the brand’s existing investors. Hexa is partnering with different brands in fashion, furniture; consumer electronics, etc, and helping them turn their digital inventory into 3D renders.
What Is Hexa Offering?
- Transform 2D product pages to 3D.
- Enhance existing 3D images.
- Assess performance.
Yehiel Atias, CEO of Hexa explained that with the advancements in artificial intelligence and NVIDIA GPUs, it was the right time to go 3D. He added,
“Just a few years ago, the digital infrastructure to do this was still so expensive that it was more affordable to arrange a photographer, models, and lighting. But with the advancements of AI and NVIDIA GPUs, it’s now feasible for retailers to use synthetic data to replace physical photoshoots.”
Popular digital retailers including Amazon, Crate & Barrel and Macy’s are using Hexa’s 3D renders. The company reduces the need for physical photoshoots of every item in a retailer’s catalog by producing thousands of renders each month. Hexa’s offering is not just economical but also environmentally friendly as the brand predicts that by imaging products digitally rather than physically, it may help consumers save up to 300 pounds of carbon emissions for each product.
AI-Accelerated Renders – How Do They Work?
With cutting-edge technology, Hexa can turn a single low-quality 2D image or a collection of such images into a high-fidelity 3D object. Depending on the complexity of the shape, the quantity of visual data that needs to be recreated, and how closely the object resembles the Hexa dataset, the company employs varying degrees of automation for its renders.
The team uses a variety of AI algorithms created using the PyTorch deep learning framework and are executed on cloud-based NVIDIA Tensor Core GPUs.
To automate certain workflow components, the team uses a variety of AI algorithms that were created using the PyTorch deep learning framework and are executed on cloud-based NVIDIA Tensor Core GPUs. For example, when a Hexa artist is recreating a 3D toaster, one algorithm can locate similar geometries generated in the past and extract it for the artist as a headstart. Similarly, another neural network is capable of scanning a retailer’s website to assess the number of products Hexa can turn into 3D renders.
The company uses NVIDIA GPUs (available on Amazon Web Services) for its rendering pipeline. Segev Nahari, lead technical artist at Hexa highlighted the ease of use through AWS as it gives them the option to thousands of NVIDIA GPUs without informing beforehand. He also predicted that rendering on NVIDIA GPUs can be up to three times faster than depending on CPUs.
[To share your insights with us, please write to firstname.lastname@example.org].