Microsoft to collaborate with DJI on drone applications
Microsoft is working with Chinese drone manufacturer DJI on an SDK for Windows 10 that will allow developers to create applications for drones. Microsoft is collaborating with Qualcomm on machine vision.
In addition to offering the SDK together, DJI and Microsoft will also develop drone applications themselves, including for use in agriculture, the construction sector and public safety. The companies use Azure IoT Edge for this. This is Microsoft’s cloud service that allows companies to make internet-of-things devices ‘smarter’, including through machine learning.
Microsoft announced the partnership at the start of its Build developer event. There, the company is also announcing that it will release the Azure IoT Edge Runtime under an open source license to allow the developer community to debug, check and modify the code. In addition, Microsoft adds Azure Cognitive Services to Azure IoT Edge, which allows companies to add speech recognition, image recognition, translation, and search capabilities to IoT devices. First, the api for Custom Vision, or image recognition, comes to the service.
Microsoft is working with Qualcomm on an image recognition developer kit. The kit should make it easier for companies to combine cameras with the hardware acceleration that Qualcomm’s Vision Intelligence Platform and AI Engine offer when processing image information. Microsoft’s part is that these companies can join the Azure Machine Learning service, for visual training based on large amounts of data.
As an extension of this is Microsoft’s introduction of Project Kinect for Azure. This is a package of sensors, including a new version of the Kinect depth camera, that developers can link to Azure services. The intention is that Azure calculations will help with the precise determination of, for example, hand movements or the placement of objects in a room.
Not only in the field of image, but for sound Microsoft had announcements. For example, there is a Speech Devices SDK, which enables speech recognition, even if there are multiple audio sources. According to Microsoft, developers can use these for drive-thru ordering services, among other things. For the Azure Cognitive Services there is now a universal Speech service, for not only speech recognition but also text-to-speech and translations.