LearningModelBinding Class

Reference

Definition

Namespace:: Windows.AI.MachineLearning

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

Used to bind values to named input and output features.

public ref class LearningModelBinding sealed : IIterable<IKeyValuePair<Platform::String ^, Platform::Object ^> ^>, IMapView<Platform::String ^, Platform::Object ^>

/// [Windows.Foundation.Metadata.Activatable(Windows.AI.MachineLearning.ILearningModelBindingFactory, 65536, "Windows.AI.MachineLearning.MachineLearningContract")]
/// [Windows.Foundation.Metadata.ContractVersion(Windows.AI.MachineLearning.MachineLearningContract, 65536)]
/// [Windows.Foundation.Metadata.MarshalingBehavior(Windows.Foundation.Metadata.MarshalingType.Agile)]
/// [Windows.Foundation.Metadata.Threading(Windows.Foundation.Metadata.ThreadingModel.Both)]
class LearningModelBinding final : IIterable<IKeyValuePair<winrt::hstring, IInspectable const&>>, IMapView<winrt::hstring, IInspectable const&>

[Windows.Foundation.Metadata.Activatable(typeof(Windows.AI.MachineLearning.ILearningModelBindingFactory), 65536, "Windows.AI.MachineLearning.MachineLearningContract")]
[Windows.Foundation.Metadata.ContractVersion(typeof(Windows.AI.MachineLearning.MachineLearningContract), 65536)]
[Windows.Foundation.Metadata.MarshalingBehavior(Windows.Foundation.Metadata.MarshalingType.Agile)]
[Windows.Foundation.Metadata.Threading(Windows.Foundation.Metadata.ThreadingModel.Both)]
public sealed class LearningModelBinding : IEnumerable<KeyValuePair<string,object>>, IReadOnlyDictionary<string,object>

function LearningModelBinding(session)

Public NotInheritable Class LearningModelBinding
Implements IEnumerable(Of KeyValuePair(Of String, Object)), IReadOnlyDictionary(Of String, Object)

Inheritance: Object Platform::Object IInspectable LearningModelBinding

Attributes: ActivatableAttribute ContractVersionAttribute MarshalingBehaviorAttribute ThreadingAttribute

Implements: IIterable<IKeyValuePair<K,V>> IEnumerable<KeyValuePair<K,V>> IIterable<IKeyValuePair<String,Object>> IEnumerable<KeyValuePair<String,Object>> IIterable<IKeyValuePair<Platform::String,Platform::Object>> IIterable<IKeyValuePair<winrt::hstring,IInspectable>> IMapView<String,Object> IReadOnlyDictionary<String,Object> IMapView<Platform::String,Platform::Object> IMapView<winrt::hstring,IInspectable>

Windows requirements

Device family	Windows 10, version 1809 (introduced in 10.0.17763.0)
API contract	Windows.AI.MachineLearning.MachineLearningContract (introduced in v1.0)

Examples

The following example retrieves the first input and output features, creates an output frame, binds the input and output features, and evaulates.

private async Task EvaluateModelAsync(
    VideoFrame _inputFrame, 
    LearningModelSession _session, 
    IReadOnlyList<ILearningModelFeatureDescriptor> _inputFeatures, 
    IReadOnlyList<ILearningModelFeatureDescriptor> _outputFeatures,
    LearningModel _model)
{
    ImageFeatureDescriptor _inputImageDescription;
    TensorFeatureDescriptor _outputImageDescription;
    LearningModelBinding _binding = null;
    VideoFrame _outputFrame = null;

    try
    {
        // 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

Models specify their input and output features using unique string names. Before evaluating the model you can bind your inputs and outputs to the session using those names. To do this you use the LearningModelBinding object which you can create based on a session.

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.

Constructors

LearningModelBinding(LearningModelSession)

Creates a LearningModelBinding from the specified LearningModelSession.

Properties

Size	Gets the number of elements in the map.

Methods

Bind(String, Object, IPropertySet)	Bind a value to the named feature using properties to control the binding.
Bind(String, Object)	Bind a value to the named feature.
Clear()	Remove all bindings.
First()	Returns an iterator that is initialized to the first element in the map view.
HasKey(String)	Determines whether the map view contains the specified key.
Lookup(String)	Returns the item at the specified key in the map view.
Split(IMapView<String,Object>, IMapView<String,Object>)	Splits the map view into two views.

Applies to

Jaa