Microsoft has added two new tools to its Azure AI cloud products to give developers the ability to add richer artificial intelligence capabilities to their apps and services for enterprise users.
The new tools, Azure Custom Vision and an Anomaly Detector service, were unveiled in a March 26 post on the Azure Blog by Anand Raman, the Azure AI platform product manager at Microsoft.
Azure Custom Vision, which is now in general availability from Microsoft, provides automated machine learning to quickly and accurately identify objects within images, wrote Raman, while Anomaly Detector is a new cognitive service that lets developers detect unusual patterns or rare events in their data that could translate to identifying problems like credit card fraud. Anomaly Detector is now available in a preview version.
"Powered by machine learning, Custom Vision makes it easy and fast for developers to build, deploy, and improve custom image classifiers to quickly recognize content in imagery," wrote Raman. "Developers can train their own classifier to recognize what matters most in their scenarios or export these custom classifiers to run them offline and in real time on iOS (in CoreML), Android (in TensorFlow) and many other devices on the edge."
Improvements in the newly released version of Custom Vision include higher quality models using a new machine learning back end for improved performance, especially on challenging datasets and fine-grained classification, as well as simpler integration of computer vision capabilities into applications with 3.0 REST APIs and SDKs, wrote Raman. "The end to end pipeline is designed to support the iterative improvement of models, so you can quickly train a model, prototype in real world conditions and use the resulting data to improve the model which gets models to production quality faster."
In addition, the exported models are optimized for the constraints of mobile devices, providing high throughput while still maintaining high accuracy, while now also allowing developers to export classifiers to support Azure Resource Manager (ARM) for Raspberry Pi 3 and the Vision AI Dev Kit.
The preview version of Anomaly Detector is used today by more than 200 teams across Azure and other core Microsoft products to boost the reliability of their systems by detecting irregularities in real time and accelerating troubleshooting, wrote Raman. "Through a single API, developers can easily embed anomaly detection capabilities into their applications to ensure high data accuracy and automatically [report] incidents as soon as they happen."
Common use case scenarios include identifying business incidents and text errors, monitoring internet of things (IoT) device traffic, detecting fraud, responding to changing markets and more, he wrote. "For instance, content providers can use Anomaly Detector to automatically scan video performance data specific to a customer's key performance indicators (KPIs), helping to identify problems in an instant. Alternatively, video streaming platforms can apply Anomaly Detector across millions of video data sets to track metrics. A missed second in video performance can translate to significant revenue loss for content providers that monetize on their platform."
Both new services are part of Microsoft's work to continue to improve Azure AI to help developers and data scientists deploy, manage and secure AI functions directly into their applications, wrote Raman. That work includes leveraging machine learning to build and train predictive models to improve business productivity with Azure Machine Learning; applying an AI-powered search experience and indexing technologies to glean insights with Azure Search; and building applications that integrate prebuilt and custom AI capabilities like vision, speech, language, search and knowledge, he added.
"Today's milestones illustrate our commitment to make the Azure AI platform suitable for every business scenario, with enterprise-grade tools that simplify application development and industry leading security and compliance for protecting customers' data," wrote Raman.