Export data and labels from Azure Custom Vision to azure ML

Thanks for the question, You can export your labeled images and tags from Azure Custom Vision to Azure Machine Learning by using the AzureML Python SDK and the Custom Vision Python SDK. You can use these SDKs to automate the process of exporting your data and labels from Custom Vision and importing them into Azure ML as data assets. This will allow you to use your labeled images and tags in the Azure ML Designer.

You can indeed export the labeled images from Azure Custom Vision and import them into Azure ML. Here are the steps you can follow:

1. Export Labeled Images from Azure Custom Vision: You can use the GetTaggedImages function to export labeled images from your Custom Vision project. This function is part of the Custom Vision API and allows you to retrieve all tagged images in a project.
2. Download the Images: Once you have retrieved the tagged images, you can download them using the download URIs provided by the GetTaggedImages function.
3. Upload Images to Azure Blob Storage: The images you work with in Azure Machine Learning must be available in an Azure Blob Storage datastore. If you don’t have an existing datastore, you can upload your downloaded images to a new datastore when you create a project.
4. Import Images into Azure ML: Once your images are in Azure Blob Storage, you can import them into Azure ML as a dataset. You can then use this dataset in your machine learning experiments.

Thank you,

But how do I import the labels in Azure ML ?

Thanks for the details, To import labels into Azure ML, you can follow these steps:

1. Export Labels from Source: your labels are in Azure Custom Vision, you’ll need to export them first. This usually involves calling an API function that retrieves all labels.
2. Download Labels: Once you’ve exported the labels, you can download them to your local machine.
3. Upload Labels to Azure Blob Storage: Azure ML works with data stored in Azure Blob Storage. If you don’t already have a datastore, you can create one and upload your labels.
4. Create a Dataset in Azure ML: With your labels in Azure Blob Storage, you can create a dataset in Azure ML. If your labels are in a CSV file, you can use the from_delimited_files method. If your labels are part of your file paths, you can use the from_files method.
5. Use the Dataset in Your Experiments: You can now use this dataset in your machine learning experiments.