Forum Discussion
Choose IoT deveice for testing: Vision AI Developer Kit vs Azure Kinect DK
Hi,
We would like to buy some camera in order to practice Azure IoT, AI. In general we want to develop cases when camera films and recognizes persons, safety helmets and other.
At the moment we are thinking between: Vision AI Developer Kit and Azure Kinect DK.
Could you advice which one is better to buy and what advantages would be one against other?
We think maybe is better to spend more money and to acquire Azure Kinect? will it make the same as Vision AI dev Kie?
Azure Kinect is not an "intelligent edge" camera meaning that it needs laptop or other HW to be connected and run inferencing. Vision AI Developer Kit is a perfect way to get started with Azure IoT and AI. It may fall short in some aspects when you start doing more comprehensive AI models etc.
We have a collection of information and example projects in A@Edge community pages. Take a look what you can do with different kinds of dev kits. NVidia's Jetson Nano and any Intel power dev kit are also good choices. They typically require a bit more advanced user to take the dev kit into use (but you'll also be able to do more) where as Vision AI Dev Kit allows to do a "no code" AI development when you use customvision.ai to build your AI models.
2 Replies
- Jussi NiemelaFormer Employee
Azure Kinect is not an "intelligent edge" camera meaning that it needs laptop or other HW to be connected and run inferencing. Vision AI Developer Kit is a perfect way to get started with Azure IoT and AI. It may fall short in some aspects when you start doing more comprehensive AI models etc.
We have a collection of information and example projects in A@Edge community pages. Take a look what you can do with different kinds of dev kits. NVidia's Jetson Nano and any Intel power dev kit are also good choices. They typically require a bit more advanced user to take the dev kit into use (but you'll also be able to do more) where as Vision AI Dev Kit allows to do a "no code" AI development when you use customvision.ai to build your AI models.
- jkubickaFormer Employee
Hi Vaidas Bražionis the Vision AI Dev Kit would actually be perfect for your scenario. We have a hard hat project / sample code that you can train, deploy, and tweak on the device.
Here's the project background and sample code:
https://azure.github.io/Vision-AI-DevKit-Pages/docs/community_project02/