Intersection of Awesome: Building AI enabled IoT solutions powered by the cloud
summary
Trains a classifier for a physical gauge (green/yellow/red) using Custom Vision, exports a compressed Qualcomm-compatible model, stores it in Azure Storage and deploys it as a Docker module from an Azure container registry to an edge device running the Vision AI dev kit and Azure IoT Edge. The device runs offline, hosts modules via the Moby runtime, and reports module twin state to IoT Hub. Demo classifies live video with ~66.7% precision (≈50% recall) and shows remote redeploys and latency but successful inference.
AI-written summary · about
chapters (8)
- 0:00Start
- 5:08Custom Vision
- 10:24Iot Hub
- 17:29Container Registry
- 17:54Demo
- 23:36Vision Kit
- 23:59Azure Iot Edge
- 27:51Cloud Interface
related talks
01h 02mAI on the Edge
45mFrom Data to AI on The Edge
49mEnable IoT with Edge Computing and Machine Learning
36mAI-Powered Real-Time Quality Control
01h 00mLearn Live: Get started with AI on Azure
41mAI + Machine Learning for .NET Developers
06mLightning Talk: Deploying AI Solutions on the Edge with Containers
31mBuild Your Intelligent Edge