IoT devices can work with cognitive services in the cloud for ML tasks such as face verification. However, it is often useful to have ML at the edge as well, to avoid streaming up data all the time – like a “wake word”. We will show you how Azure Sphere enables you to easily build ML at the edge that works with ML in the cloud, in the context of a face detection/recognition scenario.
Learn more reading the blog post at https://aka.ms/iotshow/MLOnAzureSphere
01:34 Scenarios requiring running ML at the edge on Microcontrollers
03:25 Demo scenario and architecture
08:50 Use cases for using local ML in combination with Cloud ML
12:28 Security and privacy with Azure Sphere
14:58 Wrap up and call to action