Merge pull request #897 from rnett/rn_sd_examples

[WIP] SameDiff examples

Author: AlexDBlack

dl4j-examples/src/main/java/org/deeplearning4j/examples/samediff/tfimport/SameDiffTFImportMobileNetExample.java

@@ -0,0 +1,127 @@
+package org.deeplearning4j.examples.samediff.tfimport;
+
+import java.io.File;
+import java.io.IOException;
+import java.net.URL;
+import java.util.Arrays;
+import org.apache.commons.io.FilenameUtils;
+import org.datavec.image.loader.ImageLoader;
+import org.deeplearning4j.zoo.model.helper.InceptionResNetHelper;
+import org.deeplearning4j.zoo.util.imagenet.ImageNetLabels;
+import org.nd4j.autodiff.samediff.SameDiff;
+import org.nd4j.linalg.api.ndarray.INDArray;
+import org.nd4j.linalg.api.ops.DynamicCustomOp;
+import org.nd4j.linalg.factory.Nd4j;
+import org.nd4j.linalg.indexing.INDArrayIndex;
+import org.nd4j.linalg.indexing.NDArrayIndex;
+import org.nd4j.resources.Downloader;
+
+/**
+ * This example shows the ability to import and use Tensorflow models, specifically mobilenet, and use them for inference.
+ */
+public class SameDiffTFImportMobileNetExample {
+
+    public static void main(String[] args) throws Exception {
+
+        // download and extract a tensorflow frozen model file (usually a .pb file)
+        File modelFile = downloadModel();
+
+        // import the frozen model into a SameDiff instance
+        SameDiff sd = SameDiff.importFrozenTF(modelFile);
+
+        System.out.println(sd.summary());
+
+        System.out.println("\n\n");
+
+        // get the image from https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/img/image2.jpg for testing
+        INDArray testImage = getTestImage();
+
+        // preprocess image with inception preprocessing
+        INDArray preprocessedImage = inceptionPreprocessing(testImage, 224, 224);
+
+        // Input and output names are found by looking at sd.summary() (printed earlyer).
+        // The input variable is the output of no ops, and the output variable is the input of no ops.
+
+        // Alternatively, you can use sd.outputs() and sd.inputs().
+
+        System.out.println("Input: " + sd.inputs());
+        System.out.println("Output: " + sd.outputs());
+
+        // Do inference for a single batch.
+        INDArray out = sd.batchOutput()
+            .input("input", preprocessedImage)
+            .output("MobilenetV2/Predictions/Reshape_1")
+            .execSingle();
+
+        // ignore label 0 (the background label)
+        out = out.get(NDArrayIndex.all(), NDArrayIndex.interval(1, 1001));
+
+        // get the readable label for the classes
+        String label = new ImageNetLabels().decodePredictions(out);
+
+        System.out.println("Predictions: " + label);
+
+    }
+
+
+
+    public static String MODEL_URL = "https://storage.googleapis.com/mobilenet_v2/checkpoints/mobilenet_v2_1.0_224.tgz";
+
+    // download and extract the model file in the ~/dl4j-examples-data directory used by other examples
+    public static File downloadModel() throws Exception{
+        String dataDir = FilenameUtils.concat(System.getProperty("user.home"), "dl4j-examples-data/tf_resnet");
+        String modelFile = FilenameUtils.concat(dataDir, "mobilenet_v2_1.0_224.tgz");
+
+        File frozenFile = new File(FilenameUtils.concat(dataDir, "mobilenet_v2_1.0_224_frozen.pb"));
+
+        if(frozenFile.exists()){
+            return frozenFile;
+        }
+
+        Downloader.downloadAndExtract("tf_resnet", new URL(MODEL_URL), new File(modelFile), new File(dataDir), "519bba7052fd279c66d2a28dc3f51f46", 5);
+
+        return frozenFile;
+    }
+
+    // gets the image we use to test the network.
+    // This isn't a single class ImageNet image, so it won't do very well, but it will at least classify it as a dog or a cat.
+    public static INDArray getTestImage() throws IOException {
+        URL url = new URL("https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/img/image2.jpg?raw=true");
+        return new ImageLoader(358, 500, 3).asMatrix(url.openStream());
+    }
+
+    /**
+     * Does inception preprocessing.  Takes an image with shape [c, h, w]
+     * and returns an image with shape [1, height, width, c].
+     *
+     * Eventually this will be made part of DL4J.
+     *
+     * @param height the height to resize to
+     * @param width the width to resize to
+     */
+    public static INDArray inceptionPreprocessing(INDArray img, int height, int width){
+        // add batch dimension
+        img = Nd4j.expandDims(img, 0);
+
+        // change to channels-last
+        img = img.permute(0, 2, 3, 1);
+
+        // normalize to 0-1
+        img = img.div(256);
+
+        // resize
+        INDArray preprocessedImage = Nd4j.createUninitialized(1, height, width, 3);
+
+        DynamicCustomOp op = DynamicCustomOp.builder("resize_bilinear")
+            .addInputs(img)
+            .addOutputs(preprocessedImage)
+            .addIntegerArguments(height, width).build();
+        Nd4j.exec(op);
+
+        // finish preprocessing
+        preprocessedImage = preprocessedImage.sub(0.5);
+        preprocessedImage = preprocessedImage.mul(2);
+        return preprocessedImage;
+    }
+
+}