All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.openimaj.image.processing.face.recognition.FaceRecognitionEngine Maven / Gradle / Ivy

Go to download

Implementation of a flexible face-recognition pipeline, including pluggable detectors, aligners, feature extractors and recognisers.

There is a newer version: 1.3.10
Show newest version
/**
 * Copyright (c) 2011, The University of Southampton and the individual contributors.
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 *   * 	Redistributions of source code must retain the above copyright notice,
 * 	this list of conditions and the following disclaimer.
 *
 *   *	Redistributions in binary form must reproduce the above copyright notice,
 * 	this list of conditions and the following disclaimer in the documentation
 * 	and/or other materials provided with the distribution.
 *
 *   *	Neither the name of the University of Southampton nor the names of its
 * 	contributors may be used to endorse or promote products derived from this
 * 	software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
 * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */
package org.openimaj.image.processing.face.recognition;

import java.io.DataInput;
import java.io.DataOutput;
import java.io.File;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import java.util.Set;

import org.apache.log4j.Logger;
import org.openimaj.data.dataset.GroupedDataset;
import org.openimaj.data.dataset.ListDataset;
import org.openimaj.feature.FeatureExtractor;
import org.openimaj.image.FImage;
import org.openimaj.image.processing.face.detection.DatasetFaceDetector;
import org.openimaj.image.processing.face.detection.DetectedFace;
import org.openimaj.image.processing.face.detection.FaceDetector;
import org.openimaj.io.IOUtils;
import org.openimaj.io.ReadWriteableBinary;
import org.openimaj.ml.annotation.AnnotatedObject;
import org.openimaj.ml.annotation.ScoredAnnotation;
import org.openimaj.util.pair.IndependentPair;

/**
 * The {@link FaceRecognitionEngine} ties together the implementations of a
 * {@link FaceDetector} and {@link FaceRecogniser}, and provides a single
 * convenience API with which to interact with a face recognition system.
 * 
 * @author Jonathon Hare ([email protected])
 * 
 * @param 
 *            Type of {@link DetectedFace}
 * @param 
 *            Type representing a person
 */
public class FaceRecognitionEngine
		implements
		ReadWriteableBinary
{
	private static final Logger logger = Logger.getLogger(FaceRecognitionEngine.class);

	protected FaceDetector detector;
	protected FaceRecogniser recogniser;

	protected FaceRecognitionEngine() {
	}

	/**
	 * Construct a {@link FaceRecognitionEngine} with the given face detector
	 * and recogniser.
	 * 
	 * @param detector
	 *            the face detector
	 * @param recogniser
	 *            the face recogniser
	 */
	public FaceRecognitionEngine(final FaceDetector detector,
			final FaceRecogniser recogniser)
	{
		this.detector = detector;
		this.recogniser = recogniser;
	}

	/**
	 * Create a {@link FaceRecognitionEngine} with the given face detector and
	 * recogniser.
	 * 
	 * @param 
	 *            Type of {@link DetectedFace}
	 * @param 
	 *            Type of {@link FeatureExtractor}
	 * @param 
	 *            Type representing a person
	 * 
	 * @param detector
	 *            the face detector
	 * @param recogniser
	 *            the face recogniser
	 * @return new {@link FaceRecognitionEngine}
	 */
	public static , PERSON>
			FaceRecognitionEngine
			create(final FaceDetector detector, final FaceRecogniser recogniser)
	{
		return new FaceRecognitionEngine(detector, recogniser);
	}

	/**
	 * @return the detector
	 */
	public FaceDetector getDetector() {
		return this.detector;
	}

	/**
	 * @return the recogniser
	 */
	public FaceRecogniser getRecogniser() {
		return this.recogniser;
	}

	/**
	 * Save the {@link FaceRecognitionEngine} to a file, including all the
	 * internal state of the recogniser, etc.
	 * 
	 * @param file
	 *            the file to save to
	 * @throws IOException
	 *             if an error occurs when writing
	 */
	public void save(final File file) throws IOException {
		IOUtils.writeBinaryFull(file, this);
	}

	/**
	 * Load a {@link FaceRecognitionEngine} previously saved by
	 * {@link #save(File)}.
	 * 
	 * @param 
	 *            Type of {@link DetectedFace}
	 * @param 

* Type representing a person * * @param file * the file to read from * @return the created recognition engine * @throws IOException * if an error occurs during the read */ public static FaceRecognitionEngine load(final File file) throws IOException { final FaceRecognitionEngine engine = IOUtils.read(file); return engine; } /** * Train with a dataset * * @param dataset * the dataset */ public void train(final GroupedDataset, FImage> dataset) { final GroupedDataset, FACE> faceDataset = DatasetFaceDetector .process(dataset, this.detector); this.recogniser.train(faceDataset); } /** * Train the recogniser with a single example, returning the detected face. * If multiple faces are found, the biggest is chosen. *

* If you need more control, consider calling {@link #getDetector()} to get * a detector which you can apply to your image and {@link #getRecogniser()} * to get the recogniser which you can train with the detections directly. * * @param person * the person * @param image * the image with the persons face * @return the detected face */ public FACE train(final PERSON person, final FImage image) { final List faces = this.detector.detectFaces(image); if (faces == null || faces.size() == 0) { FaceRecognitionEngine.logger.warn("no face detected"); return null; } else if (faces.size() == 1) { this.recogniser.train(AnnotatedObject.create(faces.get(0), person)); return faces.get(0); } else { FaceRecognitionEngine.logger.warn("More than one face found. Choosing biggest."); final FACE face = DatasetFaceDetector.getBiggest(faces); this.recogniser.train(AnnotatedObject.create(face, person)); return face; } } /** * Train for the given face patch without doing any face detection. It is * assumed that the given image will be a cropped/aligned image of the face * as is necessary for the given recogniser. * * @param face * The detected face implementation * @param person * The person to whom this face belongs * @return The face image */ public FACE train(final FACE face, final PERSON person) { this.recogniser.train(AnnotatedObject.create(face, person)); return face; } /** * Detect and recognise the faces in the given image, returning a list of * potential people for each face. * * @param image * the image * @return a list of faces and recognitions */ public List>>> recognise(final FImage image) { final List detectedFaces = this.detector.detectFaces(image); final List>>> results = new ArrayList>>>(); for (final FACE df : detectedFaces) { results.add(new IndependentPair>>(df, this.recogniser.annotate(df))); } return results; } /** * Detect and recognise the faces in the given image, returning the most * likely person for each face. * * @param image * the image * @return a list of faces with the most likely person */ public List>> recogniseBest(final FImage image) { final List detectedFaces = this.detector.detectFaces(image); final List>> results = new ArrayList>>(); for (final FACE df : detectedFaces) { results.add(new IndependentPair>(df, this.recogniser.annotateBest(df))); } return results; } /** * Detect and recognise the faces in the given image, returning a list of * potential people for each face. The recognised people will be restricted * to the given set. * * @param image * the image * @param restrict * set of people to restrict to * @return a list of faces and recognitions */ public List>>> recognise(final FImage image, final Set restrict) { final List detectedFaces = this.detector.detectFaces(image); final List>>> results = new ArrayList>>>(); for (final FACE df : detectedFaces) { results.add(new IndependentPair>>(df, this.recogniser.annotate(df, restrict))); } return results; } /** * Detect and recognise the faces in the given image, returning the most * likely person for each face. The recognised people will be restricted to * the given set. * * @param image * the image * @param restrict * set of people to restrict to * @return a list of faces with the most likely person */ public List>> recogniseBest(final FImage image, final Set restrict) { final List detectedFaces = this.detector.detectFaces(image); final List>> results = new ArrayList>>(); for (final FACE df : detectedFaces) { results.add(new IndependentPair>(df, this.recogniser .annotateBest(df, restrict))); } return results; } @Override public void readBinary(final DataInput in) throws IOException { final String detectorClass = in.readUTF(); this.detector = IOUtils.newInstance(detectorClass); this.detector.readBinary(in); final String recogniserClass = in.readUTF(); this.recogniser = IOUtils.newInstance(recogniserClass); this.recogniser.readBinary(in); } @Override public byte[] binaryHeader() { return "FaRE".getBytes(); } @Override public void writeBinary(final DataOutput out) throws IOException { out.writeUTF(this.detector.getClass().getName()); this.detector.writeBinary(out); out.writeUTF(this.recogniser.getClass().getName()); this.recogniser.writeBinary(out); } }





© 2015 - 2024 Weber Informatics LLC | Privacy Policy