Intel® AI Skills

Published: 10/25/2019, Last Updated: 10/25/2019

Introduction

Add AI features using comprehensive optimized libraries for faster computation on CPUs, GPUs, VPUs, and Gaussian Network Accelerators (GNAs).

  • Develop inspiring applications with Intel® AI Skills libraries and samples.
  • Deliver hardware-accelerated AI applications with the included pretrained models, pre- and post-processing capabilities, and inferencing.
  • Access the out-of-the box API to enable hardware acceleration with Intel accelerators.

Downloads

Download Samples
Download Microsoft Windows Skills* Using NuGet
Get Windows Skills Samples on GitHub*

Overview

Through a standard and simple API, Intel AI Skills provides all the required components, including pretrained models, along with preprocessing, inferencing, and post-processing capabilities.

End User License Agreement

architecture diagram of Intel AI Skills

How It Works

Using the API defined by the Windows Skills* framework, implement Intel AI Skills using a plug-in to the framework. These skills abstract complexities into a single programming paradigm with standardized primitives.

  • Applications can switch AI accelerators (such as CPUs, GPUs, VPUs, and GNAs) to allow dynamic compute decisions at run time.
  • Use with operating system interfaces and assets (such as image primitives from cameras, photos, and videos) and in conjunction with non-skills-based libraries.
  • Create components within an application, just like a recipe, to address a complex scenario, or bundle the skills into a single application package.
  • Implement using C#, C++, Windows Runtime (WinRT), and .NET on a Universal Windows Platform (UWP) and desktop Windows applications.

Included Skills

Use an image or video frame as input to apply any of the skills listed.

Skill Name Description
Background Blur Segment out individuals while blurring the background image to highlight the individuals in the foreground.
Background Replacement Segment out individuals while replacing the background with a user-selected image.
Face Detection Create bounding boxes for detected faces and other attributes, such as eyes, mouths, or nose tips.
Person Detection Create bounding boxes for detected individuals.
Super Resolution Convert a low-resolution image or video frame (320 x 240 or 640 x 360) to a high-resolution image (1280 x 960 or 1280 x 720).

Sample Applications

Intel provides sample application source code for each skill to demonstrate how to use them in applications. These UWP applications are designed to take inputs directly from camera feeds or image files and render the output on the application to quickly demonstrate the functionality of the skills.

Product and Performance Information

1

Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Notice revision #20110804