Windows.AI.MachineLearning Namespace

Important

Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.

Edit

Important

For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018.

Enables apps to load machine learning models, bind features, and evaluate the results.

Classes

ImageFeatureDescriptor	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Describes the properties of the image the model is expecting.
ImageFeatureValue	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Describes the properties of the image used to pass into a model.
LearningModel	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Represents a trained machine learning model.
LearningModelBinding	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Used to bind values to named input and output features.
LearningModelDevice	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. The device used to evaluate the machine learning model.
LearningModelEvaluationResult	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Get the results of the evaluation.
LearningModelSession	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Used to evaluate machine learning models.
LearningModelSessionOptions	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Describes inference options that are used during the creation of LearningModelSession objects.
MapFeatureDescriptor	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A map is a collection of (key, value) pairs.
SequenceFeatureDescriptor	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A sequence is an array of elements.
TensorBoolean	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A boolean tensor object.
TensorDouble	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 64-bit float tensor object.
TensorFeatureDescriptor	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Tensors are multi-dimensional arrays of values.
TensorFloat	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 32-bit float tensor object.
TensorFloat16Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 16-bit float tensor object.
TensorInt16Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 16-bit signed integer tensor object.
TensorInt32Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 32-bit signed integer tensor object.
TensorInt64Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 64-bit signed integer tensor object.
TensorInt8Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. An 8-bit signed integer tensor object.
TensorString	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A string tensor object.
TensorUInt16Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 16-bit unsigned integer tensor object.
TensorUInt32Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 32-bit unsigned integer tensor object.
TensorUInt64Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 64-bit unsigned integer tensor object.
TensorUInt8Bit	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. A 8-bit unsigned integer tensor object.

Interfaces

ILearningModelFeatureDescriptor	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Describes the common properties that all features have.
ILearningModelFeatureValue	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. The instantiated value for a feature.
ILearningModelOperatorProvider	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Describes the operaators for a learning model.
ITensor	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Tensors are multi-dimensional values.

Enums

LearningModelDeviceKind	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Defines the list of device kinds that can evaluate a machine learning model.
LearningModelFeatureKind	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Input and output feature kinds for a machine learning model.
LearningModelPixelRange	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Defines the list of image nominal pixel range suppored by Windows ML. The proper value is specified in a machine learning model's metadata.
TensorKind	Important For the latest documentation about Windows Machine Learning, see What is Windows ML. That documentation describes APIs that are in the Microsoft.Windows.AI.MachineLearning namespace, which ships in the Windows App SDK. Those APIs supersede the ones documented here, which are in the Windows.AI.MachineLearning namespace, and were shipped in 2018. Defines the list of supported tensor data types.

Examples

The following example loads a model, creates an evaluation session, gets the input and output features of the model, binds those features, and evaluates.

private async Task LoadAndEvaluateModelAsync(VideoFrame _inputFrame, string _modelFileName)
{
    LearningModel _model;
    ImageFeatureDescriptor _inputImageDescription;
    TensorFeatureDescriptor _outputImageDescription;
    LearningModelBinding _binding = null;
    VideoFrame _outputFrame = null;
    LearningModelSession _session;

    try
    {
        // Load and create the model
        var modelFile = 
            await StorageFile.GetFileFromApplicationUriAsync(new Uri($"ms-appx:///Assets/{_modelFileName}"));
        _model = await LearningModel.LoadFromStorageFileAsync(modelFile);

        // Create the evaluation session with the model
        _session = new LearningModelSession(_model);

        //Get input and output features of the model
        List<ILearningModelFeatureDescriptor> inputFeatures = _model.InputFeatures.ToList();
        List<ILearningModelFeatureDescriptor> outputFeatures = _model.OutputFeatures.ToList();

        // Retrieve the first input feature which is an image
        _inputImageDescription =
                inputFeatures.FirstOrDefault(feature => feature.Kind == LearningModelFeatureKind.Image)
                as ImageFeatureDescriptor;

        // Retrieve the first output feature which is a tensor
        _outputImageDescription =
                        outputFeatures.FirstOrDefault(feature => feature.Kind == LearningModelFeatureKind.Tensor)
                        as TensorFeatureDescriptor;

        //Create output frame based on expected image width and height
        _outputFrame = new VideoFrame(
            BitmapPixelFormat.Bgra8, 
            (int)_inputImageDescription.Width, 
            (int)_inputImageDescription.Height);

        //Create binding and then bind input/output features
        _binding = new LearningModelBinding(_session);

        _binding.Bind(_inputImageDescription.Name, _inputFrame);
        _binding.Bind(_outputImageDescription.Name, _outputFrame);

        //Evaluate and get the results
        var results = await _session.EvaluateAsync(_binding, "test");
    }
    catch (Exception ex)
    {
        StatusBlock.Text = $"error: {ex.Message}";
        _model = null;
    }
}

Remarks

Windows Server

To use this API on Windows Server, you must use Windows Server 2019 with Desktop Experience.

Thread safety

This API is thread-safe.

Feedback

Was this page helpful?