Edge-side AI with interconnection equipment
Nowadays, the convergence trend of overlay technology makes these all possible if these are the typical examples of edge-to-edge artificial intelligence combined with interconnection equipment, which will not only solve many practical difficulties facing consumers and enterprises but will also greatly Improve people's quality of life and work.
Edge computing - that is, applications, data and services at the edge of the network rather than centralized data centers - will grow by 35% annually by 2023 with an estimated industrial size of 34 billion U.S. dollars by 2023. At the same time, The powerful computing power in today's systems, the development of human-inspired AI systems, and the deployment of artificial intelligence technologies outside the data center are all huge business opportunities in the marketplace.
Smart Home fully embodies the benefits of edge-based AI, and many advanced technologies can help people liberate themselves from everyday domestic chores.
For example, if you run a small company at home and send dozens of parcels a month, you have a security camera installed in your home, but how many people come and go every day, how do you protect your home? In one Real smart home, the entire residential interconnection equipment can feel and respond to any activities inside and outside the house at any time, aware of what activities are normal, what is abnormal, which activities need to take corresponding measures, etc. Smart home devices All need the ability to have deep learning at the edge to make these real-time decisions and the right judgments.
AWS DeepLens: A fully programmable, wireless video camera that supports deep learning
At the recent AWS re: Invent global summit, Intel and Amazon announced the launch of AWS DeepLens, the company's first wireless video camera designed for developers that is fully programmable and deeply tuned to support different levels of technology Of developers provide tools to design and develop artificial intelligence (AI) and machine learning products to support the development of edge-wise AI.From the data center and cloud, to the edge of embedded systems, Intel has been committed to developing a variety of products To simplify the complex process of developing AI solutions.The dynamic nature and rapid expansion of artificial intelligence workloads require a suite of self-adapting, optimized software and services to help developers build their own solutions.
Naveen Rao, Intel vice president and general manager of AI Products, said:
Intel believes that AI is capable of solving some of the world's biggest challenges and Intel is delighted to work with AWS in the area of edge and cloud-based solutions, which covers the entire AI paradigm. AWS DeepLens represents Intel and Intel The latest AI solutions developed by Amazon include the Intel® Voice Development Kit, the Amazon Echo Show and Echo Look with Intel Technology, and the Intel® Xeon® Extensible Processor for AWS's latest C5 series. The C5 series Delivers the lowest vCPU costs in the Amazon EC2 Series and is ideal for running advanced computationally intensive workloads, including machine learning and deep learning. The Intel AI Portfolio has a wide range of edge-to-cloud product and software solutions that enable developers Be focused on your own science and drive the next wave of innovation in AI.
DeepLens combines powerful processing power with an easy-to-learn user interface to enable developers to train and deploy models in the cloud DeepLens leverages Intel optimized X4 processors and embedded graphics that support object detection and recognition Depth learning software tools and libraries, including the Intel Deep Neural Network Compute Libraries, Intel clDNN, allow real-time computer vision models to run directly on the device, reducing costs and improving real-time responsiveness. Developers leverage pre-configured Good framework to start designing and creating artificial intelligence and machine learning products in minutes, DeepLens is now ready to support Apache MXNet * and will support Tensorflow * and Caffe2 * in the first quarter of 2018.
Artificial intelligence and machine learning will give birth to a new generation of smart industries, including smart home, smart retail, smart industries, etc., through intelligent interconnection with devices to make life easier for humans. Miles Kingston, general manager of Intel's smart home sector, said:
We see that technological advancements in AI and machine learning are driving a whole new round of innovation in the entire Smart Home space. DeepLens brings together Intel's expertise in hardware and software, giving developers a powerful tool to create A variety of new experiences for smart home integration has brought unlimited potential.