Get up and running in minutes, regardless of your current skill level with vision machine learning. Connect your camera to Azure IoT Hub that controls the network traffic between the device and the cloud, and see the camera in action by running a default Vision AI module that recognizes 183 different objects.
- New to Vision ML? Start building a vision model by uploading and tagging pictures, letting Azure Custom Vision Service do the heavy lifting.
- Experienced with vision ML? Use Jupyter notebooks and Visual Studio Code to create and train custom vision models using Azure Machine Learning (AML). AML services enable you to prepare data and train models. You can then convert the trained model to the custom DLC format and package it into an IoT Edge module to deploy to the Vision AI Dev Kit.
Azure IoT Hub can push your containerized vision ML models and other modules to the Vision AI DevKit with ease, whether the camera is on your desk or in another country.
Join the Community
Build the intelligent edge
As an Intelligent Edge device, the Vision AI DevKit does inferences and runs containerized Azure services locally in the device. Moving these workloads to the edge of the network means vision ML inferencing work requires less cloud interaction while also enabling quick reaction to local events, allowing operation during extended offline periods.
An Azure IoT starter kit, the Vision AI DevKit can be used with models built and trained using the Azure Machine Learning service and CustomVision.ai.
The Vision AI DevKit features the Qualcomm Visual Intelligence Platform for hardware acceleration of AI models to deliver superior inferencing performance.