More smarts for biometrics at the edge as Skymind Konduit, Nvidia partner
Biometric technology development is set to accelerate as chip design and edge computing concepts converge to help unleash advances in the biometrics industry.
Skymind Konduit, a software developer focused on artificial intelligence, is bringing its expertise to Nvidia’s Inception startup accelerator program with the aim of helping developers bring new applications to cameras, smartphones and other devices. The timely addition of money and expertise to AI development for edge devices could result in advances around improved accuracy for facial recognition, thermal imaging and other biometric identification measures.
Skymind Konduit is home to the experts who developed Eclipse Deeplearning4j, an open-source, distributed deep-learning library written for Java and Scala. The software enables developers to use a variety of deep learning algorithms on GPUs. Nvidia is a chip design company that has popularized the use of GPUs for compute intensive tasks such as AI using processors that are more efficient than the standard Intel x86 CPU architecture.
Nvidia is among a group of companies that has been advancing technologies that support AI applications in servers and devices that reside outside the traditional large data centers.
These chips are increasingly being used in edge computing devices and services for applications such as video surveillance systems that can process vast amounts of information for increased accuracy for object detection (faces being a primary example).
The Skymind Konduit team will help bring tools and technology to startups in the Nvidia program. The tools are already well-established for developing applications to run in resource constrained edge devices. Nvidia, for its part, continues to advance what those devices can do with each new generation of GPUs for their Jetson Nano line of processor modules.
Skymind Konduit is the software division of Skymind Global Ventures, an investment fund created earlier in 2020 to advance AI startups with its access to $800 million in capital.
Adam Gibson, the original developer of Deeplearning4j, heads up Konduit, giving Nvidia’s startup program access to hard-to-find AI expertise.