Yesterday, Intel officially announced its Open Visual Inference & Neural Network Optimization (OpenVINO) toolkit, formerly known as “Intel Computer Vision SDK.” This is another step in Intel’s journey into learning neural networks and open APIs. The announcement was met with very good reception across all of the IT business.
Joining Intel’s Vision Products portfolio and based on convolutional neural networks (CNN), the toolkit will help developers “streamline deep learning inference and deployment of high-performance computer vision solutions across a wide range of vertical use cases” at the edge, according to the announcement.
Tom Lantzsch, senior vice president and general manager of the Internet of Things (IoT) Group at Intel; OpenVINO release post:
With the addition of the OpenVINO toolkit to the Intel Vision Product lineup, Intel’s vision solution provides the capability to distribute AI solutions from the edge to the network to the cloud across a diverse set of products. This empowers our customers with the flexibility to economically distribute vision solutions for actionable business insights.”
Included in the toolkit are three new APIs: The Deep Learning Deployment toolkit, a common deep learning inference toolkit that scales across Intel Vision Products, and optimized functions for OpenCV* and OpenVX*. The supported software frameworks include TensorFlow, MXNet and Caffe.
The toolkit is available to download for free and to help jump-start deployments, two OpenVINO vision accelerator kits (pre-integrated offerings supporting OpenVINO to enable accelerated prototyping and deployment) are being offered by Intel and its ecosystem partners, IEI & AAEON.