Together with the new Azure IoT Tools for VS Code extension pack, we also overhauled the experience of Azure IoT Device Workbench. And there is another exciting sample contributed by Chris Lovett (Github) from Microsoft Research about running Embedded Learning Library (ELL) on MXChip IoT DevKit to achieve the keyword spotting locally.
Experience improvements for Azure IoT Device Workbench
Recently we optimized the command structure by flattening all IoT Device Workbench commands and re-organized them. The change was to make it easier for developer discover and use the command without digging into the second level as we previously designed. And it added the flexibility to add commands for our upcoming features in the Workbench.
And there is new logo design to make it more consistent with the other IoT extensions that conformed with color guidelines:
Keyword Spotting on IoT DevKit with ELL
By using the Embedded Learning Library (ELL) from Microsoft Research, now you can do voice recognition locally on your IoT DevKit without using a cloud service. It records your voice, detects what you said from a fixed list of 30 keywords, and shows the result on the DevKit screen. This is similar to the "wake words" that you might have already got used to with those smart device such as Alexa or Google Home. Only now it can be accomplished on the constraint devices like the DevKit that uses ARM Cortex M4 chip.
To try out this new sample project:
- Open VS Code with either Azure IoT Tools extension pack or Azure IoT Device Workbench extension installed.
- Press F1, type and select "Azure IoT Device Workbench: Open Examples...".
- Choose "IoT DevKit" as device, you will see all the samples. Scroll down to the very bottom of the page, you can see the sample with tutorials and source code.
- Follow the tutorial to build your own sample.
For more details about ELL, this Channel 9 video is strongly recommended:
Suggestions and Feedback
Azure IoT Device Workbench project is open-sourced on Github. If you have any feature request or encounter any issues during your daily usage, don't hesitate to create issue on our Github repository. We are all ears.