Examples of com.github.neuralnetworks.calculation.neuronfunctions.AparapiConv2D

Package com.github.neuralnetworks.calculation.neuronfunctions

Examples of com.github.neuralnetworks.calculation.neuronfunctions.AparapiConv2D

com.github.neuralnetworks.calculation.neuronfunctions.AparapiConv2D
Base class for convolutional operations (2d) This connection accept as input a single training example (as opposed to the weighted sum which works with multiple). !!! IMPORTANT !!! Aparapi only works one-dimensional arrays of primitive data types can only call member methods of the Kernel class itself.


  Matrix i1 = new Matrix(new float[] { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 }, 1);


  Matrix o = new Matrix(4, 1);


  AparapiConv2D conv = new AparapiConv2DFF(c, 1);


  conv.calculate(c, i1, o);


  // most simple case
  assertEquals(164, o.get(0, 0), 0);
  assertEquals(184, o.get(0, 1), 0);
  assertEquals(224, o.get(0, 2), 0);

View Full Code Here


  Matrix i1 = new Matrix(new float[] { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 }, 1);


  Matrix o = new Matrix(8, 1);


  AparapiConv2D conv = new AparapiConv2DFF(c, 1);


  conv.calculate(c, i1, o);


  assertEquals(164, o.get(0, 0), 0);
  assertEquals(184, o.get(0, 1), 0);
  assertEquals(224, o.get(0, 2), 0);
  assertEquals(244, o.get(0, 3), 0);

View Full Code Here

  TensorIterator it = vp.get(c.getInputLayer()).iterator();
  for (int i = 0; i < vp.get(c.getInputLayer()).getSize(); i++) {
      vp.get(c.getInputLayer()).getElements()[it.next()] = i + 1;
  }


  AparapiConv2D conv = new AparapiConv2DFF(c, vp, c.getOutputLayer());
  conv.calculate(c, vp, c.getOutputLayer());


  // most simple case
  Tensor o = vp.get(c.getOutputLayer());


  assertEquals(164, o.get(0, 0, 0, 0), 0);

View Full Code Here

  TensorIterator it = vp.get(c.getInputLayer()).iterator();
  for (int i = 0; i < vp.get(c.getInputLayer()).getSize(); i++) {
      vp.get(c.getInputLayer()).getElements()[it.next()] = i + 1;
  }


  AparapiConv2D conv = new AparapiConv2DFF(c, vp, c.getOutputLayer());
  conv.calculate(c, vp, c.getOutputLayer());


  Tensor o = vp.get(c.getOutputLayer());


  assertEquals(164, o.get(0, 0, 0, 0), 0);
  assertEquals(184, o.get(0, 0, 1, 0), 0);

View Full Code Here

TOP

Related Classes of com.github.neuralnetworks.calculation.neuronfunctions.AparapiConv2D

com.github.neuralnetworks.tensor.Tensor

com.github.neuralnetworks.test.CNNTest

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.