In this video, we take a look at using Cognitive Toolkit (CNTK)—a unified deep learning toolkit from Microsoft.

The demo included in this video was part of our Ignite talk on cloud innovation with Azure Government. We use CNTK for an image detection problem: Identifying objects within the refrigerator.  Image detection is a harder class of problem than image classification, as image detection goes beyond classification to include localization of object(s) within an image. This is the reason for dropping down into the deep learning library. (Earlier in this presentation, Steve Michelotti showed the use of Cognitive API for image classification.)

AIS developed a prototype that highlights the features and capabilities of open standards for geospatial processing and real-time data sharing through web applications. If you haven’t already, please click here to read part one

After getting the VIIRS data into our application using GeoServer, our next objective was to enhance the prototype to demonstrate some of the exciting things AIS is able to do through the use of various web technologies. Our goal was to provide a highly collaborative environment where clients on a variety of devices could all interact in real time with map data.

Figure 1: 3D Map Displaying WMS Layers

Read More…

The Windows Azure team reached out to AIS to build a Modern App proof of concept to be deployed across Microsoft Envisioning Centers worldwide. Modern App is a demo application showing Business-to-Consumer (B2C), Business-to-Business (B2B) and Enterprise Application Integration (EAI) capabilities for cloud and on-premises-hosted applications. It’s a hybrid application, consisting of the latest Microsoft enterprise technologies.

This demo app was built to show how the Microsoft stack can seamlessly integrate and help solve complex business problems. One of the main goals of this application is to show how disparate business data can easily flow from one system to another in the enterprise and/or between partner systems. Read More…