Azure Percept sample AI models

  • Artikel

Important

Retirement of Azure Percept DK:

Update 22 February 2023: A firmware update for the Percept DK Vision and Audio accessory components (also known as Vision and Audio SOM) is now available here, and will enable the accessory components to continue functioning beyond the retirement date.

The Azure Percept public preview will be evolving to support new edge device platforms and developer experiences. As part of this evolution the Azure Percept DK and Audio Accessory and associated supporting Azure services for the Percept DK will be retired March 30th, 2023.

Effective March 30th, 2023, the Azure Percept DK and Audio Accessory will no longer be supported by any Azure services including Azure Percept Studio, OS updates, containers updates, view web stream, and Custom Vision integration. Microsoft will no longer provide customer success support and any associated supporting services. For more information, please visit the Retirement Notice Blog Post.

Azure Percept enables you to develop and deploy AI models directly to your Azure Percept DK from Azure Percept Studio. Model deployment utilizes Azure IoT Hub and Azure IoT Edge.

Sample AI models

Azure Percept Studio contains sample models for the following applications:

  • people detection
  • vehicle detection
  • general object detection
  • products-on-shelf detection

With pre-trained models, no coding or training data collection is required. Simply deploy your desired model to your Azure Percept DK from the portal and open your devkit’s video stream to see the model inferencing in action. Model inferencing telemetry can also be accessed through the Azure IoT Explorer tool.

Reference solutions

A people counting reference solution is also available. This reference solution is an open-source AI application providing edge-based people counting with user-defined zone entry/exit events. Video and AI output from the on-premises edge device is egressed to Azure Data Lake, with the user interface running as an Azure website. AI inferencing is provided by an open-source AI model for people detection.

Spatial analytics pre-built solution gif.

Custom no-code solutions

Through Azure Percept Studio, you can develop custom vision and speech solutions, no coding required.

For custom vision solutions, both object detection and classification AI models are available. Simply upload and tag your training images, which can be taken directly with the Azure Percept Vision SoM of the Azure Percept DK if desired. Model training and evaluation are easily performed in Custom Vision, which is part of Azure Cognitive Services.


For custom speech solutions, voice assistant templates are currently available for the following applications:

  • Hospitality: hotel room equipped with voice-controlled smart devices.
  • Healthcare: care facility equipped with voice-controlled smart devices.
  • Inventory: inventory hub equipped with voice-controlled smart devices.
  • Automotive: automotive hub equipped with voice-controlled smart devices.

Pre-built voice assistant keywords and commands are available directly through the portal. Custom keywords and commands may be created and trained in Speech Studio, which is also part of Azure Cognitive Services.

Advanced development

Please see the Azure Percept DK advanced development GitHub for up-to-date guidance, tutorials, and examples for things like:

  • Deploying a custom AI model to your Azure Percept DK
  • Updating a supported model with transfer learning
  • And more