Microsoft launches machine learning computer vision toolkit
A preview of Windows Vision Skills has been launched by Microsoft to help developers implement and integrate machine learning and computer vision solutions such as facial biometrics, the company announced in a blog post.
Windows Vision Skills is a set of APIs that work as Windows Runtime Components for performing photo and video analysis tasks with artificial intelligence. Prebuilt skills for object detection, skeleton detection, and emotion recognition are available at launch. The Windows Vision Skills framework standardizes the way computer vision modules fit into a Windows application as they run on the local device, according to a post by Microsoft Developer Writer Eliot Cowley.
Benefits of the Skills include simple integration, abstracting hardware acceleration, interoperability, and NuGet packages that ease iteration and updates while preserving intellectual property through licensing, as well as extensibility and modularity. Vision Skills can be added to any UWP, .NET Core, or Win32 application through a WinRT API.
Other tech giants have also launched toolkits for computer vision, including Intel’s OpenVINO, which Honeywell is using to integrate biometric facial recognition and other capabilities into its MAXPRO security platform.