Azure机器学习分类

Question

Is there an algorithm I can use to train Azure ML to recognize images. Let's say I would like to flag any picture in my collection that contains a car. Is there a way to do that with Azure ML set of algorithm and tools?

Thanks

Answer 1

You can use the ProjectOxford APIs which use ML to, amongst other things, categorize an image . You would specifically look for a trans_car tagged image (see this example )

Answer 2

You can find a pre-trained image classifier module in the OpenCV Library which creates a image classification model for 'frontal faces' using the OpenCV Library. This is all available in Azure ML right now. But the documentation says more will be available in the future.

Documentation is here - if it helps you tweak this module for your use: https://msdn.microsoft.com/library/azure/403eb35a-83b2-4191-a212-7716292c7a5b