On Wednesday, the Intel introduced a toolkit called OpenVINO, the toolkit is designed to facilitate the application of computer vision and deep learning reasoning capabilities to edge computing.
The OpenVINO (Open Visual Reasoning and Neural Network Optimization) toolkit enables developers to build and train artificial intelligence models on the cloud (such as TensorFlow, MXNet, and Caffe, and other popular frameworks) and deploy them to a variety of products.
Last year, the Intel launched the Movidius Myriad X vision processor (VPU) which was a low power SOC processor for a neural computing engine, hoping to inject AI into everyday devices through its chips. The launch of OpenVINO will further promote Intel’s layout on edge computing.
OpenVINO provides a set of optimizations and a runtime engine that can take full advantage of Intel’s technology on different artificial intelligence accelerators, allowing developers to run their models on the architecture that best suits their needs, whether in the CPU, FPGA, or Movidius VPUs are other processors.
For example, in a retail environment, developers may wish to deploy computer vision capabilities in a range of edge applications such as point-of-sale, digital signage, and security cameras, which OpenVINO can easily implement.
Intel says, with the development of AI technology and people’s attention to it, the application scene of computer vision is no longer limited to industry. With the gradual development of low-power CV chips and the development of supporting development tools, real AI is getting closer and closer.