Voxel51 secures $30M to help gen AI understand visual input more accurately


Join us in returning to NYC on June 5th to collaborate with executive leaders in exploring comprehensive methods for auditing AI models regarding bias, performance, and ethical compliance across diverse organizations. Find out how you can attend here.


Voxel51 has raised $30 million in new funding to develop its visual AI platform, which is designed to reduce the failure rate of AI projects. The company says this is because models haven’t been refined to interpret the data properly. Voxel51 wants to solve this by helping developers create more accurate and robust production-ready AI that can be used within their applications.

The new investment was led by Bessemer Venture Partners with participation from Tru Arrow Partners, Drive Capital, Top Harvest Capital, Shasta Ventures and ID Ventures. Voxel51 says it’ll use the funds to add support for new data modalities, support the next level of dataset scale, and deepen its technology integrations. In addition, the company plans to hire more in its sales, marketing and support teams—it currently has more than 30 employees—, build out its AI research science team, and invest more in its open-source contributions.

“No one would build and tune a race car on a grass field and expect to win a Formula 1 race,” Voxel Chief Executive and co-founder Brian Moore said in a statement. “Yet, organizations building visual AI are all too often forced to build and fine-tune models with mislabeled, inadequate and under-representative data and then hope for success in production. We’re helping our customers and community build better AI applications by bringing their models and data together in one place…”

voxel51 lidar screenshot
Image credit: Voxel51

LG Electronics, Berkshire Grey, Precision Planting, RIOS Intelligent Machines and Forsight are some of the enterprise organizations that use Voxel51’s products, which include its popular open-source project FiftyOne and an enterprise version called FiftyOne Teams. The technology is claimed to have improved team productivity by up to 50 percent and model accuracy by up to 30 percent.

VB Event

The AI Impact Tour: The AI Audit

Join us as we return to NYC on June 5th to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.

Request an invite

As a computer vision company helping businesses optimize unstructured data, Voxel51 is seeing interest from many industries, whether agriculture, aviation, driving, healthcare, manufacturing, retail or security. It’s not far-fetched to believe its tech is being used to help self-driving cars better understand their environment, farmers use the platform to assess their crops, or medical professionals provide better patient diagnoses.

“Voxel51 enables AI builders to work together to develop models and applications using visual data faster and more easily, making it possible for them to deliver innovative AI applications that raise product quality, increase safety, improve efficiency and create better experiences for customers,” the company explains.

It’s a multimodal AI world these days, and organizations need to think intelligently when building out their models. Not taking care of the datasets can lead to accusations of model bias and hallucinations, reducing the usefulness of the AI. Things get more complex when factoring in that most data traffic isn’t text-based but images and videos.

The $30 million investment comes nearly two years after Voxel51 raised its Series A round. To date, it has received nearly $46 million in total funding. The company declined to reveal its valuation.



Source link

About The Author

Scroll to Top