Edge AI technologies-seen and unseen- provide huge benefits in a world that's digitally connected, 24x7. This rapid advancement of edge AI is not because of one or two "killer apps"-new solutions and usages continue to emerge all the time.
Artificial Intelligence (AI) continues to show up in our everyday lives, but its presence is gentle and welcomed, largely due to the advancements in edge AI. Many AI use cases are best suited for the edge where processing happens at or close to the data source, lowering costs, reducing application or service latency, improving reliability and increasing data privacy. The use cases below deliver some of the biggest benefits of edge AI in our everyday lives, with many deployed using Intel's technologies and strong developer ecosystem.
Edge AI technologies make all this possible. As we continue to build ever-stronger data privacy policies and security technologies, it's important to focus on how edge AI contributes to our quality of life. Software is at the heart of building blocks of technology. Listed below are some of the software tools, frameworks and libraries that help developers to build applications to bring the benefits of edge AI into our lives.
Use your preferred deep learning framework on any number of hardware architectures, for both training and inference with the open source C++ library and runtime/compiler suite, nGraph Library. Whether you prefer Chainer*, MXNet*, BigDL, Theano*, Caffe*, TensorFlow*, or neonTM.
Accelerate computer vision and deep learning algorithm development across multiple types of Intel® platforms with the Intel® Distribution of OpenVINOTM toolkit. It includes a deep learning deployment toolkit with a model optimizer and inference engine, more than 30 pre-trained models (and supports 100+ public/custom models), computer vision algorithms, and samples. It also includes traditional computer vision functions like OpenCV, OpenVX*, media encode/decode, and more.
Boost performance, power efficiency, and reliability for system and IoT applications; speed-up and consolidate workloads; move from prototype to production quickly with an all-in-one foundational - tool suite, Intel® System Studio. It includes a compiler, performance libraries, analyzer and debugger tools, along with advanced cloud connectors and provides access to 400+ sensors.