Package ai.djl.basicmodelzoo.cv.object_detection.yolo

Class YOLOV3.Builder

java.lang.Object
ai.djl.basicmodelzoo.cv.object_detection.yolo.YOLOV3.Builder
Enclosing class:
YOLOV3
public static final class YOLOV3.Builder extends Object
The Builder to construct a YOLOV3 object.

  • Constructor Details

    • Builder

      public Builder()

  • Method Details

    • setNumClasses

      public YOLOV3.Builder setNumClasses(int numClasses)
      Sets the number of classes for yolov3.
      Parameters:
      numClasses - the number of classes
      Returns:
      this Builder

    • optBatchNormMomentum

      public YOLOV3.Builder optBatchNormMomentum(float batchNormMomentum)
      Sets the momentum for batchNorm layer.
      Parameters:
      batchNormMomentum - the momentum for batchNorm layer
      Returns:
      this Builder

    • optLeakyAlpha

      public YOLOV3.Builder optLeakyAlpha(float leakyAlpha)
      Sets the alpha for leakyRelu activation.
      Parameters:
      leakyAlpha - the alpha for leakyRelu activation
      Returns:
      this Builder

    • optDarkNetOutSize

      public YOLOV3.Builder optDarkNetOutSize(int darkNetOutSize)
      Sets the out size of darkNet for testing.
      Parameters:
      darkNetOutSize - the out size of darkNet
      Returns:
      this Builder

    • build

      public ai.djl.nn.Block build()
      Builds a YOLOV3 block.
      Returns:
      a YOLOV3 block

    • buildDarkNet

      public ai.djl.nn.Block buildDarkNet()
      Builds a Block that represents the backbone of yolov3, which is called DarkNet53. This can be used for testing and transfer learning.
      Returns:
      a Block that represents darkNet53