Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
/*******************************************************************************
* Copyright (c) 2015-2018 Skymind, Inc.
*
* This program and the accompanying materials are made available under the
* terms of the Apache License, Version 2.0 which is available at
* https://www.apache.org/licenses/LICENSE-2.0.
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations
* under the License.
*
* SPDX-License-Identifier: Apache-2.0
******************************************************************************/
package org.deeplearning4j.nn.layers.objdetect;
import lombok.Data;
import org.nd4j.linalg.api.ndarray.INDArray;
/**
* A detected object, by an object detection algorithm.
* Note that the dimensions (for center X/Y, width/height) depend on the specific implementation.
* For example, in the {@link Yolo2OutputLayer}, the dimensions are grid cell units - for example, with 416x416 input,
* 32x downsampling, we have 13x13 grid cells (each corresponding to 32 pixels in the input image). Thus, a centerX
* of 5.5 would be xPixels=5.5x32 = 176 pixels from left. Widths and heights are similar: in this example, a with of 13
* would be the entire image (416 pixels), and a height of 6.5 would be 6.5/13 = 0.5 of the image (208 pixels).
*
* @author Alex Black
*/
@Data
public class DetectedObject {
private final int exampleNumber;
private final double centerX;
private final double centerY;
private final double width;
private final double height;
private final INDArray classPredictions;
private int predictedClass = -1;
private final double confidence;
/**
* @param exampleNumber Index of the example in the current minibatch. For single images, this is always 0
* @param centerX Center X position of the detected object
* @param centerY Center Y position of the detected object
* @param width Width of the detected object
* @param height Height of the detected object
* @param classPredictions Row vector of class probabilities for the detected object
*/
public DetectedObject(int exampleNumber, double centerX, double centerY, double width, double height,
INDArray classPredictions, double confidence){
this.exampleNumber = exampleNumber;
this.centerX = centerX;
this.centerY = centerY;
this.width = width;
this.height = height;
this.classPredictions = classPredictions;
this.confidence = confidence;
}
/**
* Get the top left X/Y coordinates of the detected object
*
* @return Array of length 2 - top left X and Y
*/
public double[] getTopLeftXY(){
return new double[]{ centerX - width / 2.0, centerY - height / 2.0};
}
/**
* Get the bottom right X/Y coordinates of the detected object
*
* @return Array of length 2 - bottom right X and Y
*/
public double[] getBottomRightXY(){
return new double[]{ centerX + width / 2.0, centerY + height / 2.0};
}
/**
* Get the index of the predicted class (based on maximum predicted probability)
* @return Index of the predicted class (0 to nClasses - 1)
*/
public int getPredictedClass(){
if(predictedClass == -1){
if(classPredictions.rank() == 1){
predictedClass = classPredictions.argMax().getInt(0);
} else {
// ravel in case we get a column vector, or rank 2 row vector, etc
predictedClass = classPredictions.ravel().argMax().getInt(0);
}
}
return predictedClass;
}
public String toString() {
return "DetectedObject(exampleNumber=" + exampleNumber + ", centerX=" + centerX + ", centerY=" + centerY +
", width=" + width + ", height=" + height + ", confidence=" + confidence
+ ", classPredictions=" + classPredictions + ", predictedClass=" + getPredictedClass() + ")";
}
}
