Element14 Sells Google AIY Project Kits
Coral USB Accelerator Provides ML Processing


On the 22nd, Element14 began selling the 'Google AIY Project Kit' and 'Coral USB Accelerator', which simplify the development of intelligent devices capable of neural network inference.

“Most engineers struggle to know where to start when applying AI to their projects,” said Lee Turner, Global Head of Semiconductor and SBC at element14. “We invested in the Google AIY Kit and Coral USB Accelerator to bridge this knowledge gap.”

“Element14 customers can now quickly take their creative on-device local AI application ideas from prototyping to launch by adding voice and image recognition capabilities to their projects through Google’s technology,” he said.

The 'Google AIY Voice Kit (G950-00865-01)' allows users to build their own natural language processor and connect it to the 'Google Assistant' to experiment with AI and machine learning. The kit allows users to add speech recognition and AI processing to their Raspberry Pi projects.

Users can apply sample code and the Google Cloud Speech-to-Text service to convert voice commands into text and trigger program code actions. Projects that incorporate voice recognition can use key phrase detection to control robots, music, games, and more.

The Google AIY Vision Kit (G950-00866-01) includes all the components and software needed to experiment with image recognition using neural networks. Users can create their own intelligent camera capable of viewing and recognizing up to 1,000 common objects and detecting faces, emotions, and poses. It can also perform object segmentation using advanced image detection modeling. Powered by a Raspberry Pi, the kit runs deep neural networks in real-time on the device, enabling computer vision without a cloud connection.

The Coral USB Accelerator (G950-01456-01) is an easy-to-build and deploy accessory that provides high-precision, custom image classification for intelligent devices powered by AutoML Vision Edge. Users can connect the Google Edge TPU coprocessor to existing systems via a USB port, enabling high-speed machine learning inference across a wide range of systems. The onboard Edge TPU is capable of 4 trillion operations per second (TOPS), enabling advanced mobile vision models at approximately 400 frames per second.