News

Microsoft Upgrades Cognitive Services Amid AI Cloud Wars

• By David Ramel
• 03/26/2019

In the war for artificial intelligence cloud dominance with the likes of Amazon AI and Google AI, Microsoft's Azure AI team today fired off a salvo that boosts its Cognitive Services offering.

As enterprises seek to leverage AI to transform their businesses, gain competitive advantage and so on, they are increasingly adopting cloud services for an easy on-ramp, with the advantages of scalability, agility and more.

Microsoft's latest rearming of its arsenal boosts Cognitive Services, just one part of the Azure AI platform, which also includes: knowledge mining with Azure Search; several machine learning services; and AI apps and agents, wherein lies Cognitive Services.

The Azure AI team today (March 26) announced new anomaly detection functionality and Custom Vision for identifying objects in images with Azure Cognitive Services.

Microsoft says Cognitive Services help developers use intelligent algorithms in apps, Web sites and bots "to see, hear, speak, understand and interpret your user needs through natural methods of communication."

On the "see" front, a new Anomaly Detector helps developers -- as the name suggests -- detect irregularities such as unusual patterns in data or rare events, which Microsoft said can helps users foresee problems before they occur.

Detecting data anomalies serves a number of use cases, helping with things like:

• Identifying fraud, such as credit card misuse
• Identifying business incidents and text errors
• Monitoring Internet of Things (IoT) device traffic
• Responding to changing markets

The other main enhancement highlighted by Microsoft is the general availability of Custom Vision to help identify objects in images.

New improvements to Custom Vision include:

• High quality models: Custom Vision features advanced training with a new machine learning back-end for improved performance, especially on challenging datasets and fine-grained classification.
• Iterate with ease: Custom Vision helps developers integrate computer vision capabilities into applications with 3.0 REST APIs and SDKs.
• Train in the cloud, run anywhere: The exported models are optimized for the constraints of a mobile device, providing incredible throughput while still maintaining high accuracy. Now, developers can also export classifiers to support Azure Resource Manager (ARM) for Raspberry Pi 3 and the Vision AI Dev Kit.

The Custom Vision release notes list many more new features.

Microsoft said more than 1 million developers have tried Cognitive Services. "From using speech recognition, translation, and text-to-speech to image and object detection, Azure Cognitive Services makes it easy for developers to add intelligent capabilities to their applications in any scenario," the team said in an announcement blog post today (March 26).

Microsoft has been quite busy on the cloud AI front recently, having updated its ML.NET machine learning offering along with updating a new machine learning platform for Apache Spark.

Microsoft's competitors haven't been standing still, either, as Amazon Web Services (AWS) boosted its own AI services on several fronts recently.

Meanwhile, Google recently open sourced a library for training large-scale neural network models. Also, Google last July added more AI options, as CIO reported.