Ceva and embedUR methods have partnered to launch the ModelNova platform for Ceva’s NeuPro NPU household, geared toward increasing the sting AI mannequin ecosystem.
ModelNova gives developer-ready AI fashions for ultra-low-power NeuPro-Nano NPUs and high-performance NeuPro-M NPUs with augmented wide-ranging purposes akin to wearables, sensible house units, robotics and AR/VR.
The platform comes with pre-built fashions, datasets and use-case blueprints thus accelerating the time to prototype and deploy AI options.
“ModelNova is a strong extension of our NeuPro ecosystem, giving builders entry to a various and rising library of optimized AI fashions,” says Elia Shenberger, senior director of product advertising, AI Division at Ceva. “Our partnership with embedUR permits prospects to construct smarter, extra responsive edge units with minimal growth overhead.”
The fashions are deployed and optimized utilizing TensorFlow Lite Micro, which is built-in with Ceva’s NeuPro Studio SDK, enabling purposes akin to object detection, face recognition and anomaly detection.
The initiative focuses on “bodily AI,“ which mainly injects intelligence proper into the system to make it function autonomously with out a lot dependence on cloud connectivity.
Ceva’s extremely scalable NeuPro NPU household is tailor-made to satisfy the ultra-low energy and high-performance necessities in a broad vary of edge AI purposes, from entry-level to high-end linked units.
The ModelNova providing is straight away out there and can add new fashions and use instances on a rolling foundation.
Earlier this 12 months, Ceva introduced a partnership with ALi Company integrating Ceva’s NPUs into ALi’s Video Show Sub-System (VDSS) platform.
Associated
AI/ML | Ceva | system intelligence | edge AI | edge computing | embedded methods | embedUR | NPU
