GANs.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/jansoe/AICA/blob/main/GANs.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RX02S6Aq8OxF"
      },
      "source": [
        "# Generative Adversial Networks\n",
        "\n",
        "GANs, bei denen ein *generatives* Netz versucht, ein *diskriminatives* Netz auszutricksen."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IKgoIQgiDQ0G",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "43f87f45-3b53-48cd-bf8d-707f7b534b39"
      },
      "source": [
        "import tensorflow as tf\n",
        "# Es lohnt sich zu überprüfen, ob man auch wirklich eine GPU bekommen hat.\n",
        "print(\"Anzahl der zugewiesenen GPUs:\", len(tf.config.list_physical_devices('GPU')))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Anzahl der zugewiesenen GPUs: 1\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c6sSUrDuDXkX"
      },
      "source": [
        "import time\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from tensorflow.keras import layers\n",
        "from IPython import display"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "8j_Db4FUKZ10"
      },
      "source": [
        "Mit den Autoencodern haben Sie Ihr bereits eine Art von generativem Netz erstellt. Eine weitere Klasse von generativen Algorithmen sind die sogenannten *Generative Adversarial Networks*. \n",
        "\n",
        "Bei dieser Technik handelt es sich um zwei Netzwerke, die gegeneinander ausgespielt werden. Ein *generatives* Netzwerk erstellt künstliche Bilder, während ein *diskriminatives* Netzwerk zu klassifizieren versucht, ob ein Bild aus einem Trainingsset von *echten* Bildern stammt oder vom generativen Netzwerk erzeugt wurde. Die beiden Netzwerke werden dabei gleichzeitig trainiert und schaukeln sich gegenseitig hoch.\n",
        "\n",
        "Ein typisches Beispiel aus der Popkultur sind [künstlich generierte, fotorealistische Bilder von Menschen, die es nicht gibt](https://this-person-does-not-exist.com/en).\n",
        "\n",
        "### Datensatz für ein Anwendungsbeispiel\n",
        "\n",
        "Das Trainieren von GANs auf echten Bilddaten ist sehr rechenintensiv und auch komplizierter. Zur Illustration des Prinzips benutzen wir deshalb wieder den MNIST-Datensatz."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xuSzCk3RaDM4"
      },
      "source": [
        "# Herunterladen des Datensatzes\n",
        "(train_images, train_labels), (test_images, test_labels) = tf.keras.datasets.mnist.load_data()\n",
        "# Hinzufügen einer weiteren Dimension für Convolutional Networks\n",
        "train_images = train_images[:,:,:,np.newaxis].astype('float32')\n",
        "\n",
        "# Normalize the images to [-1, 1]\n",
        "train_images_gan = (train_images - 127.5) / 127.5"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RoDc3cGrDm-V"
      },
      "source": [
        "# Hier erstellen wir ein Objekt, das die Portionierung der Daten während des Trainings handhabt.\n",
        "train_dataset = tf.data.Dataset.from_tensor_slices(train_images_gan)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zeGa3cnHdQkf"
      },
      "source": [
        "### Das Generator Modell\n",
        "\n",
        "Analog zum Generator im Autoencoder erstellen wir hier auch wieder ein Modell, das aus einem kompakten Eingabecode Ausgabebilder erzeugt. Die Anzahl der Dimensionen der zugehörigen latenten Variable setzen wir diesmal jedoch auf 100."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QZ6CMvSdG-hL"
      },
      "source": [
        "noise_dim = 8\n",
        "\n",
        "generator = tf.keras.Sequential()\n",
        "\n",
        "generator.add(layers.Dense(7*7*256, use_bias=False, input_shape=(noise_dim,)))\n",
        "generator.add(layers.BatchNormalization())\n",
        "generator.add(layers.LeakyReLU())\n",
        "\n",
        "generator.add(layers.Reshape((7, 7, 256)))\n",
        "\n",
        "generator.add(layers.Conv2DTranspose(128, kernel_size=(5, 5), strides=1, padding='same', use_bias=False))\n",
        "generator.add(layers.BatchNormalization())\n",
        "generator.add(layers.LeakyReLU())\n",
        "\n",
        "generator.add(layers.Conv2DTranspose(64, kernel_size=(5, 5), strides=2, padding='same', use_bias=False))\n",
        "generator.add(layers.BatchNormalization())\n",
        "generator.add(layers.LeakyReLU())\n",
        "\n",
        "generator.add(layers.Conv2DTranspose(1, kernel_size=(5, 5), strides=2, padding='same', use_bias=False, activation='tanh'))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wv-EvxjQLAbT",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "16dc7f48-4e5e-42c9-9963-71c7c029d86c"
      },
      "source": [
        "tf.keras.utils.plot_model(generator, show_shapes=True, dpi=80)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CVu9tnUK4i5s"
      },
      "source": [
        "Wie zuvor können wir den Generator mit Rauschen füttern. Da er noch nicht trainiert ist, bekommen wir auch nur Rauschen zurück:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cU_PnXljH1zk",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 447
        },
        "outputId": "f3ba0998-a5fe-4ae5-afb3-5d4a784804d7"
      },
      "source": [
        "np.random.seed(42)\n",
        "noise = tf.random.normal([1, noise_dim])\n",
        "generated_image = generator(noise, training=False)\n",
        "\n",
        "plt.imshow(generated_image[0, :, :, 0], cmap='gray')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f10505f2830>"
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6wM6-J0_dyEk"
      },
      "source": [
        "### Der Diskriminator\n",
        "\n",
        "Der Diskriminator ist ein einfaches Klassifikationsnetz, wie wir es bereits aus vorangegangenen Aufgaben kennen."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QlUgRR5cH5oN"
      },
      "source": [
        "discriminator = tf.keras.Sequential()\n",
        "discriminator.add(layers.Conv2D(64, (5, 5), strides=2, padding='same', input_shape=[28, 28, 1]))\n",
        "discriminator.add(layers.LeakyReLU())\n",
        "discriminator.add(layers.Dropout(0.3))\n",
        "\n",
        "discriminator.add(layers.Conv2D(128, (5, 5), strides=2, padding='same'))\n",
        "discriminator.add(layers.LeakyReLU())\n",
        "discriminator.add(layers.Dropout(0.3))\n",
        "\n",
        "discriminator.add(layers.Flatten())\n",
        "discriminator.add(layers.Dense(1))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2fVx0v7AQWbq",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 816
        },
        "outputId": "fa7d3d78-7047-4dd4-d177-7402ee9602be"
      },
      "source": [
        "tf.keras.utils.plot_model(discriminator, show_shapes=True, dpi=80)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ewj7yrbA48sy"
      },
      "source": [
        "Um zu sehen, ob wir die Dimensionen richtig hinbekommen haben, generieren wir eine Vorhersage für das eben generierte Bild."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A4ryYFfmINtH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "afd432b2-533e-4e51-bbb6-90bfcc7690e8"
      },
      "source": [
        "decision = discriminator(generated_image)\n",
        "print(decision)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor([[0.00030805]], shape=(1, 1), dtype=float32)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mlTZtV148sYr"
      },
      "source": [
        "Zur Kontrolle füttern wir dem Modell nun noch eines der echten Bilder."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "znTwQ4S08q42",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7b430751-f131-4100-a976-57a02bb4aaaf"
      },
      "source": [
        "decision = discriminator(train_images[:1])\n",
        "print(decision)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tf.Tensor([[10.670707]], shape=(1, 1), dtype=float32)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A8zOW3sY5QCA"
      },
      "source": [
        "Der Diskriminator funktioniert, ist sich aber aufgrund seiner Unerfahrenheit noch unsicher, was wir von ihm wollen."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B27XtJHMeRJw"
      },
      "source": [
        "### Fehlerdefinition\n",
        "\n",
        "Bisher haben wir zwei Netzwerke erstellt, zwischen denen noch kein Zusammenhang besteht. Dieser wird durch die Berechnung der Fehler hergestellt. Hierfür ist es hilfreich, zunächst den grundsätzlichen Ablauf des Trainings kennenzulernen.\n",
        "\n",
        "In jedem Trainingsschritt wird für ein Batch (also eine kleine Portion) von \"echten\" Bildern (Bilder des Trainingsdatensatzes) zunächt ein gleichgroßer Stapel \"falscher\" Bilder vom Generator erzeugt (`g_out`). Für beide Arten von Bildern - das \"reale\" Bild des Trainingsdatensatzes und das generierte Bild - werden Antworten des Diskriminator-Netzwerks errechnet (`d_out_real`, `d_out_gen`). Aus dieser Vorhersage wird dann für beide Netze ein Fehler und damit eine Gewichtsveränderung zur Minimierung des jeweiligen Fehlers berechnet:\n",
        "\n",
        "**Generator**:\n",
        "- `1 - d_out_gen`: Je niedriger der Diskriminator auf das generierte Bild reagiert, desto größer der Fehler des Generators.\n",
        "\n",
        "**Diskriminator**: (Der Fehler des Diskriminators ergibt sich aus zwei Werten)\n",
        "- `1 - d_out_train`: Je niedriger der Output für die Bilder des Trainingsdatensatzes, desto höher ist der Fehler.\n",
        "- `d_out_gen`: Je höher der Output für die Bilder des Generators, desto höher ist der Fehler.\n",
        "\n",
        "Da die Fehlerart von Generator und Diskriminator die gleiche wie bei einem gewöhnlichen binären Klassifikationsproblem ist, können wir diese für beide gemeinsam definieren:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A0glw4tvIgYI"
      },
      "source": [
        "# This method returns a helper function to compute cross entropy loss\n",
        "cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HMEvaRUg6XH9"
      },
      "source": [
        "Die Fehlerfunktion des Generators wird wie beschreiben auf dem Output des Diskriminators berechnet. Ziel des Generators ist es, den Diskriminator davon zu überzeugen, dass alle seine Bilder *einsen* sind:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LheY9YfdJSNL"
      },
      "source": [
        "# Der Generator soll künstliche Bilder produzieren, so dass der Diskriminator den Generator-Output mit möglichst hoher Wahrscheinlichkeit als echt klassifiziert \n",
        "def generator_loss(d_out_gen):\n",
        "    generator_loss = cross_entropy(tf.ones_like(d_out_gen), d_out_gen)\n",
        "    return generator_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "33vW-3G_avjI"
      },
      "source": [
        "Der Diskriminator soll lernen, die Wahrscheinlichkeit dafür auszugeben, dass ein Bild \"echt\", d.h. nicht generiert ist. Er reagiert also idealerweise mit einer 1 auf ein Bild des Trainingsdatensatzen und mit einer 0 auf ein vom Generator-Netzwerk generiertes Bild."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CGo-Y94oJIIl"
      },
      "source": [
        "def discriminator_loss(d_out_real, d_out_gen):\n",
        "    real_loss = cross_entropy(tf.ones_like(d_out_real), d_out_real)\n",
        "    fake_loss = cross_entropy(tf.zeros_like(d_out_gen), d_out_gen)\n",
        "    discriminator_loss = real_loss + fake_loss\n",
        "    return discriminator_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_ose4Ckb6p9e"
      },
      "source": [
        "Nun optimieren wir abwechseld den Discriminator und Generator"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yfOUJ0hoKOkS"
      },
      "source": [
        "generator_optimizer = tf.keras.optimizers.Adam(1e-4)\n",
        "discriminator_optimizer = tf.keras.optimizers.Adam(1e-4)\n",
        "\n",
        "@tf.function # mit diesem Zusatz können wir das Training etwas beschleunigen\n",
        "def train_step(images):\n",
        "    \n",
        "    noise = tf.random.normal([images.shape[0], noise_dim])\n",
        "    # Trainings des Diskriminators\n",
        "    with tf.GradientTape() as disc_tape:\n",
        "        generator_out = generator(noise, training=False)\n",
        "        discriminator_out_real = discriminator(images, training=True)\n",
        "        discriminator_out_generated = discriminator(generator_out, training=True)\n",
        "\n",
        "        disc_loss = discriminator_loss(discriminator_out_real, discriminator_out_generated)\n",
        "\n",
        "    gradients_of_discriminator = disc_tape.gradient(disc_loss, discriminator.trainable_variables)\n",
        "    discriminator_optimizer.apply_gradients(zip(gradients_of_discriminator, discriminator.trainable_variables))\n",
        "    \n",
        "    # Training des Generators\n",
        "    with tf.GradientTape() as gen_tape:\n",
        "        generator_out = generator(noise, training=True)\n",
        "        discriminator_out_generated = discriminator(generator_out, training=False)\n",
        "\n",
        "        gen_loss = generator_loss(discriminator_out_generated)\n",
        "        \n",
        "    gradients_of_generator = gen_tape.gradient(gen_loss, generator.trainable_variables)\n",
        "    generator_optimizer.apply_gradients(zip(gradients_of_generator, generator.trainable_variables))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ItpzgYOJ7e7F"
      },
      "source": [
        "Wir generieren uns einige Werte für die latente Variable, um den Trainingsfortschritt zu visualisieren: "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rNH7XwA2aOCt"
      },
      "source": [
        "# We will reuse these random codes overtime to visualize progress during training\n",
        "test_input = tf.random.normal([16, noise_dim])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_PvkzFJZK1R6",
        "cellView": "form"
      },
      "source": [
        "#@title\n",
        "#@markdown Bitte ausführen: Code zum Visualisieren des Trainings\n",
        "\n",
        "def generate_and_show_images(model, test_input):\n",
        "  # Notice `training` is set to False.\n",
        "  # This is so all layers run in inference mode\n",
        "  # dropouts: rescaled and no dropout\n",
        "  # batchnorm: frozen\n",
        "  # other layers: weights frozen\n",
        "  predictions = model(test_input, training=False)\n",
        "\n",
        "  fig = plt.figure(figsize=(4,4))\n",
        "\n",
        "  for i in range(predictions.shape[0]):\n",
        "      plt.subplot(4, 4, i+1)\n",
        "      plt.imshow(predictions[i, :, :, 0] * 127.5 + 127.5, cmap='gray')\n",
        "      plt.axis('off')\n",
        "\n",
        "  plt.show()\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h8hwd1Qw72m3"
      },
      "source": [
        "Dann beginnt die eigentliche Trainingsschleife:"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7BBO5XSCKj5R",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 362
        },
        "outputId": "b140f489-8cb5-481f-a469-0f59c1b6ea2b"
      },
      "source": [
        "train_dataset = train_dataset.shuffle(60000)\n",
        "train_dataset = train_dataset.batch(256)\n",
        "\n",
        "for epoch in range(50):\n",
        "    start = time.time()\n",
        "\n",
        "    for image_batch in train_dataset:\n",
        "        train_step(image_batch)\n",
        "\n",
        "    # Produce images as we go\n",
        "    display.clear_output(wait=True)\n",
        "    generate_and_show_images(generator, test_input)\n",
        "\n",
        "    print('Time for epoch {} is {} sec'.format(epoch + 1, time.time()-start))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 400x400 with 16 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Time for epoch 50 is 13.650860071182251 sec\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pIlTJsl4fd8g"
      },
      "source": [
        "### Ergebnis\n",
        "\n",
        "Nach einigen Iterationen erzeugt der Generator Bilder, die wie handgeschriebene Zahlen oder zumindest Symbole aussehen. Beim Autoencoder bildeten sich bestimmte Bereiche im latenten Raum heraus, an denen einzelne Ziffern repräsentiert werden. Diese Bereiche konnten wir finden, indem wir sie mit dem Encoder *aktivieren*. Beim Generator der GANs ist das nicht mehr möglich - er generiert zufällige Bilder, ohne dass wir direkten Einfluss nehmen können. \n",
        "\n",
        "Generieren Wir zum Spaß also noch einige Zufallsbilder. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sYq2A1W6JFkd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 209
        },
        "outputId": "335e20d5-2a25-4056-d0e0-388be5aa88f4"
      },
      "source": [
        "for i in range(5):\n",
        "    plt.subplot(1,5,i+1)\n",
        "    random_code = np.random.randn(1, noise_dim) # zufälliger Code!\n",
        "    image = generator.predict(random_code)\n",
        "    plt.imshow(image.squeeze(), cmap='gray')\n",
        "    plt.axis('off')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 213ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 27ms/step\n",
            "1/1 [==============================] - 0s 26ms/step\n",
            "1/1 [==============================] - 0s 25ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 5 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SdCDjvo2-IV0"
      },
      "source": [
        "### Aufgabe: Größe des latenten Raumes\n",
        "\n",
        "Im vorangegangenen Beispiel hatte unser latenter Raum eine Dimensionalität von 8 (Variable `latent_dim`), analog zum Autoencoder, der auch mit einer Code-Länge von acht Zahlen zurecht kam. Da es bei GANs nicht auf Kompression ankommt, wird hier oft ein vergleichsweise großer Raum gewählt, mit der Hoffnung, reichhaltigere Ergebnisse zu erziehlen. Verändern Sie die Variable `latent_dim` und führen Sie den Code ab dieser Zelle erneut aus und untersuchen Sie folgende Fragestellungen:\n",
        "\n",
        "- Generiert das Modell für `latent_dim=100` wesentlich bessere Ergebnisse? \n",
        "- Generiert das Modell auch noch für `latent_dim=2` vernünftige Ergebnisse?\n",
        "- Ab welcher Größe funktioniert es kaum noch?"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "dJ1IIZoxic29"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dY7MraDt-3ht"
      },
      "source": [
        "### Aufgabe: Morphing\n",
        "\n",
        "Beim Autoencoder haben Sie gesehen, wie wir aus zufälligen Codes Bilder generieren können. Kopieren Sie den Code zum Morphen zwischen Bildern des Autoencoders und modifizieren Sie ihn so, dass zwischen zwei zufällig generierten Codes des GAN-Generators gemorpht wird."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eDCklf_1ldWK"
      },
      "source": [],
      "execution_count": null,
      "outputs": []
    }
  ]
}