Interface IYoloModel
Defines an interface for a class that encapsulates a Yolo8 ONNX model for image processing tasks. This interface supports multiple tasks including prediction, classification, and segmentation. It uses the Microsoft ML framework for inference but is not thread-safe.
Inherited Members
Namespace: EyeAuras.OpenCVAuras.ML
Assembly: EyeAuras.OpenCVAuras.Shared.dll
Syntax
public interface IYoloModel : IDisposableProperties
| Name | Description |
|---|---|
| ConfidenceThreshold | Gets or sets the confidence threshold for predictions. |
| Description | Gets the description of the Yolo model, including information about its size, dimensions, and labels. |
| InferenceSession | Gets the inference session associated with the Yolo model. |
| ModelData | |
| ModelSize | Gets the size of the model's input window. |
| ModelType | Gets the type of the Yolo model indicating the kind of task it is optimized for, such as Object Detection or Segmentation. |
| OutputParser | Gets the output parser that interprets the raw outputs from the Yolo model into structured predictions. |
| OverlapThreshold | Gets or sets the overlap threshold for predictions. |
Methods
| Name | Description |
|---|---|
| Classify(Image<Bgr, byte>) | Classifies an image and returns a list of classifications. |
| Inference(DenseTensor<float>, Size) | Performs inference on a given image tensor and returns raw outputs as named ONNX values. |
| Predict(Image<Bgr, byte>, MLMaskProcessingType) | Predicts objects in an image and returns a list of predictions. Used for both object detection and segmentation tasks. |
| Predict(DenseTensor<float>, Size, MLMaskProcessingType) | Predicts objects in an image tensor and returns a list of predictions. Used for both object detection and segmentation tasks. |