Java example source code file (PreprocessorVertex.java)

This example Java source code file (PreprocessorVertex.java) is included in the alvinalexander.com "Java Source Code Warehouse" project. The intent of this project is to help you "Learn Java by Example" ^TM.

Learn more about this Java project at its project page.

Java - Java tags/keywords

cnntofeedforwardpreprocessor, cnntornnpreprocessor, feedforwardtocnnpreprocessor, feedforwardtornnpreprocessor, graphvertex, indarray, inputtype, invalidinputtypeexception, noargsconstructor, override, preprocessorvertex, rnntocnnpreprocessor, runtimeexception, string

The PreprocessorVertex.java Java example source code

/*
 *
 *  * Copyright 2016 Skymind,Inc.
 *  *
 *  *    Licensed under the Apache License, Version 2.0 (the "License");
 *  *    you may not use this file except in compliance with the License.
 *  *    You may obtain a copy of the License at
 *  *
 *  *        http://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.
 *
 */

package org.deeplearning4j.nn.conf.graph;


import lombok.Data;
import lombok.EqualsAndHashCode;
import lombok.NoArgsConstructor;
import org.deeplearning4j.nn.conf.InputPreProcessor;
import org.deeplearning4j.nn.conf.inputs.InputType;
import org.deeplearning4j.nn.conf.inputs.InvalidInputTypeException;
import org.deeplearning4j.nn.conf.preprocessor.*;
import org.deeplearning4j.nn.graph.ComputationGraph;
import org.nd4j.linalg.api.ndarray.INDArray;

/** PreprocessorVertex is a simple adaptor class that allows a {@link InputPreProcessor} to be used in a ComputationGraph
 * GraphVertex, without it being associated with a layer.
 * @author Alex Black
 */
@NoArgsConstructor
@Data @EqualsAndHashCode(callSuper=false)
public class PreprocessorVertex extends GraphVertex {

    private InputPreProcessor preProcessor;
    private InputType outputType;

    public PreprocessorVertex(InputPreProcessor preProcessor) {
        this(preProcessor, null);
    }

    /**
     * @param preProcessor The input preprocessor
     * @param outputType Override for the type of output used in {@link #getOutputType(InputType...)}. This may be necessary
     *                   for the automatic addition of other processors in the network, given a custom/non-standard InputPreProcessor
     */
    public PreprocessorVertex(InputPreProcessor preProcessor, InputType outputType) {
        this.preProcessor = preProcessor;
        this.outputType = outputType;
    }

    @Override
    public GraphVertex clone() {
        return new PreprocessorVertex(preProcessor.clone());
    }

    @Override
    public boolean equals(Object o) {
        if (!(o instanceof PreprocessorVertex)) return false;
        return ((PreprocessorVertex) o).preProcessor.equals(preProcessor);
    }

    @Override
    public int hashCode() {
        return preProcessor.hashCode();
    }

    @Override
    public int numParams(boolean backprop){
        return 0;
    }

    @Override
    public org.deeplearning4j.nn.graph.vertex.GraphVertex instantiate(ComputationGraph graph, String name, int idx,
                                                                      INDArray paramsView, boolean initializeParams) {
        return new org.deeplearning4j.nn.graph.vertex.impl.PreprocessorVertex(graph, name, idx, preProcessor);
    }

    @Override
    public InputType getOutputType(InputType... vertexInputs) throws InvalidInputTypeException {
        if (vertexInputs.length != 1) throw new InvalidInputTypeException("Invalid input: Preprocessor vertex expects "
                + "exactly one input");
        if (outputType != null) return outputType;   //Allows user to override for custom preprocessors

        //Otherwise, try to infer:
        switch (vertexInputs[0].getType()) {
            case FF:
                if (preProcessor instanceof FeedForwardToCnnPreProcessor) {
                    FeedForwardToCnnPreProcessor ffcnn = (FeedForwardToCnnPreProcessor) preProcessor;
                    return InputType.convolutional(ffcnn.getNumChannels(), ffcnn.getInputWidth(), ffcnn.getInputHeight());
                } else if (preProcessor instanceof FeedForwardToRnnPreProcessor) {
                    return InputType.recurrent(((InputType.InputTypeFeedForward)vertexInputs[0]).getSize());
                } else {
                    //Assume preprocessor doesn't change the type of activations
                    return InputType.feedForward(((InputType.InputTypeFeedForward) vertexInputs[0]).getSize());
                }
            case RNN:
                if (preProcessor instanceof RnnToCnnPreProcessor) {
                    RnnToCnnPreProcessor ffcnn = (RnnToCnnPreProcessor) preProcessor;
                    return InputType.convolutional(ffcnn.getNumChannels(), ffcnn.getInputWidth(), ffcnn.getInputHeight());
                } else if (preProcessor instanceof RnnToFeedForwardPreProcessor) {
                    return InputType.feedForward(((InputType.InputTypeRecurrent) vertexInputs[0]).getSize());
                } else {
                    //Assume preprocessor doesn't change the type of activations
                    return InputType.recurrent(((InputType.InputTypeRecurrent)vertexInputs[0]).getSize());
                }
            case CNN:
                if (preProcessor instanceof CnnToFeedForwardPreProcessor) {
                    CnnToFeedForwardPreProcessor p = (CnnToFeedForwardPreProcessor)preProcessor;
                    int outSize = p.getInputHeight()*p.getInputWidth()*p.getNumChannels();
                    return InputType.feedForward(outSize);
                } else if (preProcessor instanceof CnnToRnnPreProcessor) {
                    CnnToRnnPreProcessor p = (CnnToRnnPreProcessor)preProcessor;
                    int outSize = p.getInputHeight()*p.getInputWidth()*p.getNumChannels();
                    return InputType.recurrent(outSize);
                } else {
                    //Assume preprocessor doesn't change the type of activations
                    return vertexInputs[0];
                }
            default:
                throw new RuntimeException("Unknown InputType: " + vertexInputs[0]);
        }

    }
}