diff --git a/src/SGAN.ipynb b/src/SGAN.ipynb
index eecb18e..a648fb2 100644
--- a/src/SGAN.ipynb
+++ b/src/SGAN.ipynb
@@ -1 +1,3768 @@
-{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[]},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"widgets":{"application/vnd.jupyter.widget-state+json":{"8449be8723ae49a998eef74ac99830ae":{"model_module":"@jupyter-widgets/output","model_name":"OutputModel","model_module_version":"1.0.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/output","_model_module_version":"1.0.0","_model_name":"OutputModel","_view_count":null,"_view_module":"@jupyter-widgets/output","_view_module_version":"1.0.0","_view_name":"OutputView","layout":"IPY_MODEL_70abdf1919c642108f09f8b05e0cbbd2","msg_id":"","outputs":[{"output_type":"display_data","data":{"text/plain":"Labs... \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[35m100%\u001b[0m \u001b[36m0:00:00\u001b[0m\n","text/html":"
Labs... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00\n
\n"},"metadata":{}}]}},"70abdf1919c642108f09f8b05e0cbbd2":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"140350eb93d342db996b0ef188a99a01":{"model_module":"@jupyter-widgets/output","model_name":"OutputModel","model_module_version":"1.0.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/output","_model_module_version":"1.0.0","_model_name":"OutputModel","_view_count":null,"_view_module":"@jupyter-widgets/output","_view_module_version":"1.0.0","_view_name":"OutputView","layout":"IPY_MODEL_bb5381d4f3f54262899685bd44ad4db6","msg_id":"","outputs":[{"output_type":"display_data","data":{"text/plain":"Unlabs... \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[35m100%\u001b[0m \u001b[36m0:00:00\u001b[0m\n","text/html":"Unlabs... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00\n
\n"},"metadata":{}}]}},"bb5381d4f3f54262899685bd44ad4db6":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}},"32f04016312a486e8920415c14df928f":{"model_module":"@jupyter-widgets/output","model_name":"OutputModel","model_module_version":"1.0.0","state":{"_dom_classes":[],"_model_module":"@jupyter-widgets/output","_model_module_version":"1.0.0","_model_name":"OutputModel","_view_count":null,"_view_module":"@jupyter-widgets/output","_view_module_version":"1.0.0","_view_name":"OutputView","layout":"IPY_MODEL_00b216d0f53045c5b1533d1d4266f756","msg_id":"","outputs":[{"output_type":"display_data","data":{"text/plain":"Tests... \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[35m100%\u001b[0m \u001b[36m0:00:00\u001b[0m\n","text/html":"Tests... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00\n
\n"},"metadata":{}}]}},"00b216d0f53045c5b1533d1d4266f756":{"model_module":"@jupyter-widgets/base","model_name":"LayoutModel","model_module_version":"1.2.0","state":{"_model_module":"@jupyter-widgets/base","_model_module_version":"1.2.0","_model_name":"LayoutModel","_view_count":null,"_view_module":"@jupyter-widgets/base","_view_module_version":"1.2.0","_view_name":"LayoutView","align_content":null,"align_items":null,"align_self":null,"border":null,"bottom":null,"display":null,"flex":null,"flex_flow":null,"grid_area":null,"grid_auto_columns":null,"grid_auto_flow":null,"grid_auto_rows":null,"grid_column":null,"grid_gap":null,"grid_row":null,"grid_template_areas":null,"grid_template_columns":null,"grid_template_rows":null,"height":null,"justify_content":null,"justify_items":null,"left":null,"margin":null,"max_height":null,"max_width":null,"min_height":null,"min_width":null,"object_fit":null,"object_position":null,"order":null,"overflow":null,"overflow_x":null,"overflow_y":null,"padding":null,"right":null,"top":null,"visibility":null,"width":null}}}},"accelerator":"GPU","gpuClass":"standard"},"cells":[{"cell_type":"code","source":["!git clone https://github.com/axelcarlier/projsemisup\n","!pip install rich"],"metadata":{"id":"oip8sNddZXqQ","executionInfo":{"status":"ok","timestamp":1674483788419,"user_tz":-60,"elapsed":115397,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}},"colab":{"base_uri":"https://localhost:8080/"},"outputId":"7b061793-d20a-4ba7-d5f0-84c2d72edce4"},"execution_count":1,"outputs":[{"output_type":"stream","name":"stdout","text":["Cloning into 'projsemisup'...\n","remote: Enumerating objects: 48161, done.\u001b[K\n","remote: Total 48161 (delta 0), reused 0 (delta 0), pack-reused 48161\u001b[K\n","Receiving objects: 100% (48161/48161), 2.96 GiB | 28.92 MiB/s, done.\n","Resolving deltas: 100% (44/44), done.\n","Updating files: 100% (22857/22857), done.\n","Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Collecting rich\n"," Downloading rich-13.2.0-py3-none-any.whl (238 kB)\n","\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m238.9/238.9 KB\u001b[0m \u001b[31m6.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n","\u001b[?25hRequirement already satisfied: pygments<3.0.0,>=2.6.0 in /usr/local/lib/python3.8/dist-packages (from rich) (2.6.1)\n","Collecting markdown-it-py<3.0.0,>=2.1.0\n"," Downloading markdown_it_py-2.1.0-py3-none-any.whl (84 kB)\n","\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m84.5/84.5 KB\u001b[0m \u001b[31m12.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n","\u001b[?25hRequirement already satisfied: typing-extensions<5.0,>=4.0.0 in /usr/local/lib/python3.8/dist-packages (from rich) (4.4.0)\n","Collecting mdurl~=0.1\n"," Downloading mdurl-0.1.2-py3-none-any.whl (10.0 kB)\n","Installing collected packages: mdurl, markdown-it-py, rich\n","Successfully installed markdown-it-py-2.1.0 mdurl-0.1.2 rich-13.2.0\n"]}]},{"cell_type":"code","execution_count":2,"metadata":{"id":"qwlA5PJlI7NM","executionInfo":{"status":"ok","timestamp":1674483795463,"user_tz":-60,"elapsed":7072,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"outputs":[],"source":["import os\n","\n","from PIL.Image import open\n","from PIL.Image import ANTIALIAS\n","\n","import numpy as np\n","from numpy import expand_dims\n","from numpy import zeros\n","from numpy import ones\n","from numpy import asarray\n","from numpy.random import randn\n","from numpy.random import randint\n","\n","from keras import backend\n","from keras.models import Model\n","from keras.models import load_model\n","from keras.layers import Input\n","from keras.layers import Dense\n","from keras.layers import Reshape\n","from keras.layers import Flatten\n","from keras.layers import Conv2D\n","from keras.layers import Conv2DTranspose\n","from keras.layers import LeakyReLU\n","from keras.layers import Dropout\n","from keras.layers import Lambda\n","from keras.layers import Activation\n","from keras.layers import Concatenate\n","from keras.layers import BatchNormalization\n","from keras.metrics import SparseTopKCategoricalAccuracy\n","from keras.metrics import SparseCategoricalAccuracy\n","from keras.metrics import BinaryAccuracy\n","from keras.optimizers import Adam\n","from keras.applications import MobileNet\n","\n","import tensorflow as tf\n","import tensorflow_datasets.public_api as tfds\n","\n","import imgaug.augmenters as iaa\n","\n","from matplotlib import pyplot\n","\n","from rich.progress import track"]},{"cell_type":"code","source":["IMAGE_SIZE = 64\n","LATENT_DIM = 512\n","BATCH_SIZE = 128\n","LEARNING_RATE = 3e-5\n","\n","PATH = '/content/projsemisup/'\n","CLASSES = os.listdir(PATH + 'Lab/')\n","NB_CLASSES = len(CLASSES)\n","LAB_COUNT = len(os.listdir(PATH + 'Lab/' + CLASSES[0] + '/'))\n","TEST_COUNT = len(os.listdir(PATH + 'Test/' + CLASSES[0] + '/'))"],"metadata":{"id":"86O4uJQYY8Vd","executionInfo":{"status":"ok","timestamp":1674483944644,"user_tz":-60,"elapsed":8,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":16,"outputs":[]},{"cell_type":"code","source":["def load_semisup_data():\n","\n"," classes = os.listdir(PATH + 'Lab/')\n","\n"," # Initialise les structures de données\n"," x_lab = np.zeros((LAB_COUNT * NB_CLASSES, IMAGE_SIZE, IMAGE_SIZE, 3), dtype=np.uint8)\n"," y_lab = np.zeros((LAB_COUNT * NB_CLASSES, 1))\n"," i = 0\n"," for c in track(classes, description='Labs...'):\n","\n"," class_label = classes.index(c)\n"," list_images = os.listdir(PATH + 'Lab/' + c + '/')\n","\n"," for img_name in list_images:\n"," # Lecture de l'image\n"," img = open(PATH + 'Lab/' + c + '/' + img_name)\n","\n"," # Mise à l'échelle de l'image\n"," img = img.resize((IMAGE_SIZE, IMAGE_SIZE), ANTIALIAS)\n"," img = img.convert('RGB')\n","\n"," # Remplissage de la variable x\n"," x_lab[i] = np.asarray(img, dtype=np.uint8)\n"," y_lab[i] = class_label\n"," i = i + 1\n","\n"," list_images = os.listdir(PATH + 'Unlab/')\n"," nb_unlab = len(list_images)\n"," x_unlab = np.zeros((nb_unlab, IMAGE_SIZE, IMAGE_SIZE, 3), dtype=np.uint8)\n"," i = 0\n"," for img_name in track(list_images, description='Unlabs...'):\n"," # Lecture de l'image\n"," img = open(PATH + 'Unlab/' + img_name)\n","\n"," # Mise à l'échelle de l'image\n"," img = img.resize((IMAGE_SIZE, IMAGE_SIZE), ANTIALIAS)\n"," img = img.convert('RGB')\n","\n"," # Remplissage de la variable x\n"," x_unlab[i] = np.asarray(img, dtype=np.uint8)\n"," i = i + 1\n","\n"," file_PATH_test = os.listdir(PATH + 'Test/')\n","\n"," # Initialise les structures de données\n"," x_test = np.zeros((TEST_COUNT * NB_CLASSES, IMAGE_SIZE, IMAGE_SIZE, 3), dtype=np.uint8)\n"," y_test = np.zeros((TEST_COUNT * NB_CLASSES, 1))\n"," i = 0\n"," for c in track(file_PATH_test, description='Tests...'):\n","\n"," class_label = classes.index(c)\n"," list_images = os.listdir(PATH + 'Test/' + c + '/')\n","\n"," for img_name in list_images:\n"," # Lecture de l'image\n"," img = open(PATH + 'Test/' + c + '/' + img_name, )\n","\n"," # Mise à l'échelle de l'image\n"," img = img.resize((IMAGE_SIZE, IMAGE_SIZE), ANTIALIAS)\n"," img = img.convert('RGB')\n","\n"," # Remplissage de la variable x\n"," x_test[i] = np.asarray(img, dtype=np.uint8)\n"," y_test[i] = class_label\n"," i = i + 1\n","\n"," return (\n"," x_lab, y_lab,\n"," x_unlab,\n"," x_test, y_test\n"," )"],"metadata":{"id":"UwadDSV1Y_Sc","executionInfo":{"status":"ok","timestamp":1674483795466,"user_tz":-60,"elapsed":26,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":4,"outputs":[]},{"cell_type":"code","source":["def define_discriminator():\n","\n"," in_image = Input(shape=(IMAGE_SIZE, IMAGE_SIZE, 3))\n"," fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(in_image)\n"," fe = LeakyReLU(alpha=0.2)(fe)\n"," fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(fe)\n"," fe = LeakyReLU(alpha=0.2)(fe)\n"," fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(fe)\n"," fe = LeakyReLU(alpha=0.2)(fe)\n"," fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(fe)\n"," fe = LeakyReLU(alpha=0.2)(fe)\n"," fe = Flatten()(fe)\n"," fe = Dropout(0.4)(fe)\n"," fe = Dense(128)(fe)\n","\n"," # mbn = MobileNet(\n"," # input_shape=(IMAGE_SIZE, IMAGE_SIZE, 3),\n"," # classes=NB_CLASSES,\n"," # include_top=False,\n"," # weights=None\n"," # )\n"," # fe = Flatten()(mbn.output)\n","\n"," # Unsupervised output\n"," d_out_layer = Dense(1)(fe)\n"," d_out_layer = Activation('sigmoid')(d_out_layer)\n"," d_model = Model(in_image, d_out_layer)\n"," # d_model = Model(mbn.input, d_out_layer)\n"," opt = Adam(learning_rate=LEARNING_RATE)\n"," d_model.compile(loss='binary_crossentropy',\n"," optimizer=opt, metrics=[BinaryAccuracy()])\n","\n"," # Supervised output\n"," c_out_layer = Dense(NB_CLASSES)(fe)\n"," c_out_layer = Activation('softmax')(c_out_layer)\n"," c_model = Model(in_image, c_out_layer)\n"," # c_model = Model(mbn.input, c_out_layer)\n"," opt = Adam(learning_rate=LEARNING_RATE)\n"," c_model.compile(loss='sparse_categorical_crossentropy',\n"," optimizer=opt, metrics=[SparseCategoricalAccuracy(), SparseTopKCategoricalAccuracy(k=3)])\n","\n"," return d_model, c_model"],"metadata":{"id":"mIfaYQKnZE2T","executionInfo":{"status":"ok","timestamp":1674483795467,"user_tz":-60,"elapsed":26,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":5,"outputs":[]},{"cell_type":"code","source":["def define_generator():\n","\n"," in_lat = Input(shape=(LATENT_DIM,))\n"," gen = Dense(4*4*256)(in_lat)\n"," gen = BatchNormalization(momentum=0.8)(gen)\n"," gen = LeakyReLU(alpha=0.2)(gen)\n"," gen = Reshape((4, 4, 256))(gen)\n"," gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n"," gen = BatchNormalization(momentum=0.8)(gen)\n"," gen = LeakyReLU(alpha=0.2)(gen)\n"," gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n"," gen = BatchNormalization(momentum=0.8)(gen)\n"," gen = LeakyReLU(alpha=0.2)(gen)\n"," gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n"," gen = BatchNormalization(momentum=0.8)(gen)\n"," gen = LeakyReLU(alpha=0.2)(gen)\n"," gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n"," gen = BatchNormalization(momentum=0.8)(gen)\n"," gen = LeakyReLU(alpha=0.2)(gen)\n"," out_layer = Conv2D(3, (3, 3), padding=\"same\", activation=\"tanh\")(gen)\n","\n"," model = Model(in_lat, out_layer)\n","\n"," return model"],"metadata":{"id":"5Bcd3eRCZHkG","executionInfo":{"status":"ok","timestamp":1674483795468,"user_tz":-60,"elapsed":26,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":6,"outputs":[]},{"cell_type":"code","source":["def define_gan(g_model, d_model):\n","\n"," d_model.trainable = False\n"," gan_output = d_model(g_model.output)\n","\n"," model = Model(g_model.input, gan_output)\n","\n"," opt = Adam(learning_rate=2*LEARNING_RATE)\n"," model.compile(loss='binary_crossentropy',\n"," optimizer=opt, metrics=[BinaryAccuracy()])\n","\n"," return model"],"metadata":{"id":"0cAFqWRCZJLW","executionInfo":{"status":"ok","timestamp":1674483795469,"user_tz":-60,"elapsed":26,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":7,"outputs":[]},{"cell_type":"code","source":["def select_supervised_samples(dataset, n_samples=100):\n"," X, y = dataset\n"," X_list, y_list = list(), list()\n"," n_per_class = int(n_samples / NB_CLASSES)\n"," for i in range(NB_CLASSES):\n"," # get all images for this class\n"," X_with_class = X[y == i]\n"," # choose random instances\n"," ix = randint(0, len(X_with_class), n_per_class)\n"," # add to list\n"," [X_list.append(X_with_class[j]) for j in ix]\n"," [y_list.append(i) for j in ix]\n"," return asarray(X_list), asarray(y_list)"],"metadata":{"id":"pSDIItbJZLaR","executionInfo":{"status":"ok","timestamp":1674483795470,"user_tz":-60,"elapsed":26,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":8,"outputs":[]},{"cell_type":"code","source":["def generate_real_samples(dataset, n_samples, aug=None):\n"," # split into images and labels\n"," images, labels = dataset\n"," # choose random instances\n"," ix = randint(0, images.shape[0], n_samples)\n"," # select images and labels\n"," X = images[ix]\n"," if labels is not None:\n"," labels = labels[ix]\n"," # generate class labels\n"," y = ones((n_samples, 1))\n"," # apply augmentation\n"," if aug is not None:\n"," X = aug(images=X)\n"," # normalize batch\n"," X = X.astype(\"double\") / 127.5 - 1.0\n"," return X, labels, y"],"metadata":{"id":"x_ZiI1KdZNVT","executionInfo":{"status":"ok","timestamp":1674483795794,"user_tz":-60,"elapsed":350,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":9,"outputs":[]},{"cell_type":"code","source":["def generate_latent_points(n_samples):\n"," # generate points in the latent space\n"," z_input = randn(n_samples * LATENT_DIM)\n"," # reshape into a batch of inputs for the network\n"," z_input = z_input.reshape(n_samples, LATENT_DIM)\n"," return z_input"],"metadata":{"id":"lkUFZ6xOZPJS","executionInfo":{"status":"ok","timestamp":1674483795797,"user_tz":-60,"elapsed":30,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":10,"outputs":[]},{"cell_type":"code","source":["def generate_fake_samples(generator, n_samples):\n"," # generate points in latent space\n"," z_input = generate_latent_points(n_samples)\n"," # predict outputs\n"," images = generator.predict(z_input, verbose=0)\n"," # create class labels\n"," y = zeros((n_samples, 1))\n"," return images, y"],"metadata":{"id":"XvOWaVWsZQla","executionInfo":{"status":"ok","timestamp":1674483795799,"user_tz":-60,"elapsed":28,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":11,"outputs":[]},{"cell_type":"code","source":["def summarize_performance(step, d_model, g_model, c_model, dataset, n_samples=100):\n"," # prepare fake examples\n"," X, _ = generate_fake_samples(g_model, n_samples)\n"," # scale from [-1,1] to [0,1]\n"," X = (X + 1) / 2.0\n"," # create figure\n"," pyplot.figure(figsize=(10, 4))\n"," # plot images\n"," for i in range(10):\n"," # define subplot\n"," pyplot.subplot(2, 5, 1 + i)\n"," # turn off axis\n"," pyplot.axis('off')\n"," # plot raw pixel data\n"," pyplot.imshow(X[i, :, :, :])\n"," pyplot.show()\n"," \n"," # get test images\n"," _, _, _, x_test, y_test = dataset\n"," # normalize images\n"," x_test = x_test.astype(\"double\") / 127.5 - 1.0\n"," # evaluate the classifier model\n"," _, acc, acc3 = c_model.evaluate(x_test, y_test, verbose=0)\n"," print(f\"Test accuracy:\\n top 1: {acc*100:.3f}%\\n top 3: {acc3*100:.3f}%\")\n"," # save the discriminator model\n"," filename1 = f\"d_model_{step+1:04d}.h5\"\n"," d_model.save_weights(filename1)\n"," # save the generator model\n"," filename2 = f\"g_model_{step+1:04d}.h5\"\n"," g_model.save_weights(filename2)\n"," # save the classifier model\n"," filename3 = f\"c_model_{step+1:04d}.h5\"\n"," c_model.save_weights(filename3)\n"," print(f\"Saved: {filename1}, {filename2}, {filename3}\\n\")\n","\n"," return acc, acc3"],"metadata":{"id":"6xybda-3ZTFc","executionInfo":{"status":"ok","timestamp":1674483795801,"user_tz":-60,"elapsed":29,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":12,"outputs":[]},{"cell_type":"code","source":["def train(g_model, d_model, c_model, gan_model, dataset, aug=None, nb_epochs=100):\n"," # select supervised dataset\n"," x_lab, y_lab, x_unlab, _, _ = dataset\n"," # calculate the number of batches per training epoch\n"," bat_per_epo = int(x_unlab.shape[0] / BATCH_SIZE)\n"," # store accuracies\n"," accuracies = {'train': {'top1': [], 'top3': []}, 'val': {'top1': [], 'top3': []}}\n","\n"," print(f\"nb_epochs={nb_epochs}, batch_size={BATCH_SIZE}, b/e={bat_per_epo}\\n\")\n"," for epoch in range(nb_epochs):\n","\n"," print(f\"epoch {epoch+1:3}/{nb_epochs}:\")\n"," for step in range(bat_per_epo):\n"," # update supervised discriminator (c)\n"," X, Y, _ = generate_real_samples([x_lab, y_lab], BATCH_SIZE, aug)\n"," c_loss, c_acc, c_acc3 = c_model.train_on_batch(X, Y)\n"," # update unsupervised discriminator (d)\n"," X_real, _, Y_real = generate_real_samples([x_unlab, None], BATCH_SIZE//2)\n"," X_fake, Y_fake = generate_fake_samples(g_model, BATCH_SIZE//2)\n"," X = np.concatenate([X_real, X_fake], axis=0)\n"," Y = np.concatenate([Y_real, Y_fake], axis=0)\n"," d_loss, d_acc = d_model.train_on_batch(X, Y)\n"," # update generator (g)\n"," X_gan, y_gan = generate_latent_points(BATCH_SIZE), ones((BATCH_SIZE, 1))\n"," g_loss, g_acc = gan_model.train_on_batch(X_gan, y_gan)\n"," # show losses and accuracies\n"," print(f\"\\rstep {step+1:5}/{bat_per_epo}:\" +\n"," f\" c[{c_loss:7.3f}, {c_acc*100:3.0f}%, {c_acc3*100:3.0f}%]\" +\n"," f\" d[{d_loss:7.3f}, {d_acc*100:3.0f}%]\" +\n"," f\" g[{g_loss:7.3f}, {g_acc*100:3.0f}%]\",\n"," end='')\n"," # evaluate the model\n"," acc, acc3 = summarize_performance(epoch, d_model, g_model, c_model, dataset)\n"," accuracies['train']['top1'].append(c_acc)\n"," accuracies['train']['top3'].append(c_acc3)\n"," accuracies['val']['top1'].append(acc)\n"," accuracies['val']['top3'].append(acc3)\n"," print(accuracies)"],"metadata":{"id":"2MoC8sysY0Lc","executionInfo":{"status":"ok","timestamp":1674483795803,"user_tz":-60,"elapsed":29,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":13,"outputs":[]},{"cell_type":"code","source":["# load image data\n","dataset = load_semisup_data()"],"metadata":{"id":"_GU7lrzCaVzF","colab":{"base_uri":"https://localhost:8080/","height":65,"referenced_widgets":["8449be8723ae49a998eef74ac99830ae","70abdf1919c642108f09f8b05e0cbbd2","140350eb93d342db996b0ef188a99a01","bb5381d4f3f54262899685bd44ad4db6","32f04016312a486e8920415c14df928f","00b216d0f53045c5b1533d1d4266f756"]},"executionInfo":{"status":"ok","timestamp":1674483943912,"user_tz":-60,"elapsed":148137,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}},"outputId":"5934c561-0c6f-43c7-a9a7-dad5f29f8490"},"execution_count":14,"outputs":[{"output_type":"display_data","data":{"text/plain":["Output()"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"8449be8723ae49a998eef74ac99830ae"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":[],"text/html":["\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\n"],"text/html":["\n","
\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["Output()"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"140350eb93d342db996b0ef188a99a01"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":[],"text/html":["\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\n"],"text/html":["\n","
\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["Output()"],"application/vnd.jupyter.widget-view+json":{"version_major":2,"version_minor":0,"model_id":"32f04016312a486e8920415c14df928f"}},"metadata":{}},{"output_type":"display_data","data":{"text/plain":[],"text/html":["\n"]},"metadata":{}},{"output_type":"display_data","data":{"text/plain":["\n"],"text/html":["\n","
\n"]},"metadata":{}}]},{"cell_type":"code","source":["# augmentation\n","aug = iaa.RandAugment(n=2, m=9)"],"metadata":{"id":"vlC87HYvWUY8","executionInfo":{"status":"ok","timestamp":1674483943913,"user_tz":-60,"elapsed":15,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":15,"outputs":[]},{"cell_type":"code","source":["# create the discriminator models\n","d_model, c_model = define_discriminator()\n","# create the generator\n","g_model = define_generator()\n","# create the gan\n","gan_model = define_gan(g_model, d_model)\n","# train model\n","train(g_model, d_model, c_model, gan_model, dataset, aug)"],"metadata":{"id":"zuWf9tXqZVE-"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["backup_name = '0040'\n","\n","# create the discriminator models\n","d_model, c_model = define_discriminator()\n","# create the generator\n","g_model = define_generator()\n","\n","d_model.load_weights(f\"d_model_{backup_name}.h5\")\n","c_model.load_weights(f\"c_model_{backup_name}.h5\")\n","g_model.load_weights(f\"g_model_{backup_name}.h5\")\n","\n","# create the gan\n","gan_model = define_gan(g_model, d_model)\n","\n","# create the gan\n","gan_model = define_gan(g_model, d_model)\n","# train model\n","train(g_model, d_model, c_model, gan_model, dataset, aug)"],"metadata":{"id":"7btFJZgciMoR","colab":{"base_uri":"https://localhost:8080/","height":1000,"output_embedded_package_id":"12w46u3QlSkjLBeQWCKr_T3su-v3efvre"},"outputId":"17cc1f09-88dc-4080-ec60-fb07b874e32d","executionInfo":{"status":"ok","timestamp":1674493900342,"user_tz":-60,"elapsed":2138671,"user":{"displayName":"Damien Guillotin","userId":"16184557771337807994"}}},"execution_count":18,"outputs":[{"output_type":"display_data","data":{"text/plain":"Output hidden; open in https://colab.research.google.com to view."},"metadata":{}}]}]}
\ No newline at end of file
+{
+ "nbformat": 4,
+ "nbformat_minor": 0,
+ "metadata": {
+ "colab": {
+ "provenance": []
+ },
+ "kernelspec": {
+ "name": "python3",
+ "display_name": "Python 3"
+ },
+ "language_info": {
+ "name": "python"
+ },
+ "widgets": {
+ "application/vnd.jupyter.widget-state+json": {
+ "1b3a5e8e274742b5a04af83869731bb0": {
+ "model_module": "@jupyter-widgets/output",
+ "model_name": "OutputModel",
+ "model_module_version": "1.0.0",
+ "state": {
+ "_dom_classes": [],
+ "_model_module": "@jupyter-widgets/output",
+ "_model_module_version": "1.0.0",
+ "_model_name": "OutputModel",
+ "_view_count": null,
+ "_view_module": "@jupyter-widgets/output",
+ "_view_module_version": "1.0.0",
+ "_view_name": "OutputView",
+ "layout": "IPY_MODEL_91c8bbbabb28486f9e14f5ddda207e52",
+ "msg_id": "",
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": "Labs... \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[35m100%\u001b[0m \u001b[36m0:00:00\u001b[0m\n",
+ "text/html": "Labs... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00\n
\n"
+ },
+ "metadata": {}
+ }
+ ]
+ }
+ },
+ "91c8bbbabb28486f9e14f5ddda207e52": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "model_module_version": "1.2.0",
+ "state": {
+ "_model_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.2.0",
+ "_model_name": "LayoutModel",
+ "_view_count": null,
+ "_view_module": "@jupyter-widgets/base",
+ "_view_module_version": "1.2.0",
+ "_view_name": "LayoutView",
+ "align_content": null,
+ "align_items": null,
+ "align_self": null,
+ "border": null,
+ "bottom": null,
+ "display": null,
+ "flex": null,
+ "flex_flow": null,
+ "grid_area": null,
+ "grid_auto_columns": null,
+ "grid_auto_flow": null,
+ "grid_auto_rows": null,
+ "grid_column": null,
+ "grid_gap": null,
+ "grid_row": null,
+ "grid_template_areas": null,
+ "grid_template_columns": null,
+ "grid_template_rows": null,
+ "height": null,
+ "justify_content": null,
+ "justify_items": null,
+ "left": null,
+ "margin": null,
+ "max_height": null,
+ "max_width": null,
+ "min_height": null,
+ "min_width": null,
+ "object_fit": null,
+ "object_position": null,
+ "order": null,
+ "overflow": null,
+ "overflow_x": null,
+ "overflow_y": null,
+ "padding": null,
+ "right": null,
+ "top": null,
+ "visibility": null,
+ "width": null
+ }
+ },
+ "f4b81afd77a64eff9395edc9181c531a": {
+ "model_module": "@jupyter-widgets/output",
+ "model_name": "OutputModel",
+ "model_module_version": "1.0.0",
+ "state": {
+ "_dom_classes": [],
+ "_model_module": "@jupyter-widgets/output",
+ "_model_module_version": "1.0.0",
+ "_model_name": "OutputModel",
+ "_view_count": null,
+ "_view_module": "@jupyter-widgets/output",
+ "_view_module_version": "1.0.0",
+ "_view_name": "OutputView",
+ "layout": "IPY_MODEL_75dea27498a14b6d85cba75e3661214d",
+ "msg_id": "",
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": "Unlabs... \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[35m100%\u001b[0m \u001b[36m0:00:00\u001b[0m\n",
+ "text/html": "Unlabs... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00\n
\n"
+ },
+ "metadata": {}
+ }
+ ]
+ }
+ },
+ "75dea27498a14b6d85cba75e3661214d": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "model_module_version": "1.2.0",
+ "state": {
+ "_model_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.2.0",
+ "_model_name": "LayoutModel",
+ "_view_count": null,
+ "_view_module": "@jupyter-widgets/base",
+ "_view_module_version": "1.2.0",
+ "_view_name": "LayoutView",
+ "align_content": null,
+ "align_items": null,
+ "align_self": null,
+ "border": null,
+ "bottom": null,
+ "display": null,
+ "flex": null,
+ "flex_flow": null,
+ "grid_area": null,
+ "grid_auto_columns": null,
+ "grid_auto_flow": null,
+ "grid_auto_rows": null,
+ "grid_column": null,
+ "grid_gap": null,
+ "grid_row": null,
+ "grid_template_areas": null,
+ "grid_template_columns": null,
+ "grid_template_rows": null,
+ "height": null,
+ "justify_content": null,
+ "justify_items": null,
+ "left": null,
+ "margin": null,
+ "max_height": null,
+ "max_width": null,
+ "min_height": null,
+ "min_width": null,
+ "object_fit": null,
+ "object_position": null,
+ "order": null,
+ "overflow": null,
+ "overflow_x": null,
+ "overflow_y": null,
+ "padding": null,
+ "right": null,
+ "top": null,
+ "visibility": null,
+ "width": null
+ }
+ },
+ "c8b4fb345aa44aa7b3b5b373cab3ba9c": {
+ "model_module": "@jupyter-widgets/output",
+ "model_name": "OutputModel",
+ "model_module_version": "1.0.0",
+ "state": {
+ "_dom_classes": [],
+ "_model_module": "@jupyter-widgets/output",
+ "_model_module_version": "1.0.0",
+ "_model_name": "OutputModel",
+ "_view_count": null,
+ "_view_module": "@jupyter-widgets/output",
+ "_view_module_version": "1.0.0",
+ "_view_name": "OutputView",
+ "layout": "IPY_MODEL_93229082acb045e980e0c34a7a60ba6a",
+ "msg_id": "",
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": "Tests... \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[35m100%\u001b[0m \u001b[36m0:00:00\u001b[0m\n",
+ "text/html": "Tests... ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100% 0:00:00\n
\n"
+ },
+ "metadata": {}
+ }
+ ]
+ }
+ },
+ "93229082acb045e980e0c34a7a60ba6a": {
+ "model_module": "@jupyter-widgets/base",
+ "model_name": "LayoutModel",
+ "model_module_version": "1.2.0",
+ "state": {
+ "_model_module": "@jupyter-widgets/base",
+ "_model_module_version": "1.2.0",
+ "_model_name": "LayoutModel",
+ "_view_count": null,
+ "_view_module": "@jupyter-widgets/base",
+ "_view_module_version": "1.2.0",
+ "_view_name": "LayoutView",
+ "align_content": null,
+ "align_items": null,
+ "align_self": null,
+ "border": null,
+ "bottom": null,
+ "display": null,
+ "flex": null,
+ "flex_flow": null,
+ "grid_area": null,
+ "grid_auto_columns": null,
+ "grid_auto_flow": null,
+ "grid_auto_rows": null,
+ "grid_column": null,
+ "grid_gap": null,
+ "grid_row": null,
+ "grid_template_areas": null,
+ "grid_template_columns": null,
+ "grid_template_rows": null,
+ "height": null,
+ "justify_content": null,
+ "justify_items": null,
+ "left": null,
+ "margin": null,
+ "max_height": null,
+ "max_width": null,
+ "min_height": null,
+ "min_width": null,
+ "object_fit": null,
+ "object_position": null,
+ "order": null,
+ "overflow": null,
+ "overflow_x": null,
+ "overflow_y": null,
+ "padding": null,
+ "right": null,
+ "top": null,
+ "visibility": null,
+ "width": null
+ }
+ }
+ }
+ },
+ "accelerator": "GPU",
+ "gpuClass": "standard"
+ },
+ "cells": [
+ {
+ "cell_type": "code",
+ "source": [
+ "!git clone https://github.com/axelcarlier/projsemisup\n",
+ "!pip install rich"
+ ],
+ "metadata": {
+ "id": "oip8sNddZXqQ",
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "outputId": "28f052f6-9bb0-4cbc-8a07-d9cf996a5086"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "Cloning into 'projsemisup'...\n",
+ "remote: Enumerating objects: 48161, done.\u001b[K\n",
+ "remote: Total 48161 (delta 0), reused 0 (delta 0), pack-reused 48161\u001b[K\n",
+ "Receiving objects: 100% (48161/48161), 2.96 GiB | 30.41 MiB/s, done.\n",
+ "Resolving deltas: 100% (44/44), done.\n",
+ "Updating files: 100% (22857/22857), done.\n",
+ "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
+ "Requirement already satisfied: rich in /usr/local/lib/python3.8/dist-packages (13.2.0)\n",
+ "Requirement already satisfied: markdown-it-py<3.0.0,>=2.1.0 in /usr/local/lib/python3.8/dist-packages (from rich) (2.1.0)\n",
+ "Requirement already satisfied: pygments<3.0.0,>=2.6.0 in /usr/local/lib/python3.8/dist-packages (from rich) (2.6.1)\n",
+ "Requirement already satisfied: typing-extensions<5.0,>=4.0.0 in /usr/local/lib/python3.8/dist-packages (from rich) (4.4.0)\n",
+ "Requirement already satisfied: mdurl~=0.1 in /usr/local/lib/python3.8/dist-packages (from markdown-it-py<3.0.0,>=2.1.0->rich) (0.1.2)\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "id": "qwlA5PJlI7NM"
+ },
+ "outputs": [],
+ "source": [
+ "import os\n",
+ "\n",
+ "from PIL.Image import open\n",
+ "from PIL.Image import ANTIALIAS\n",
+ "\n",
+ "import numpy as np\n",
+ "from numpy import expand_dims\n",
+ "from numpy import zeros\n",
+ "from numpy import ones\n",
+ "from numpy import asarray\n",
+ "from numpy.random import randn\n",
+ "from numpy.random import randint\n",
+ "\n",
+ "from keras import backend\n",
+ "from keras.models import Model\n",
+ "from keras.models import load_model\n",
+ "from keras.layers import Input\n",
+ "from keras.layers import Dense\n",
+ "from keras.layers import Reshape\n",
+ "from keras.layers import Flatten\n",
+ "from keras.layers import Conv2D\n",
+ "from keras.layers import Conv2DTranspose\n",
+ "from keras.layers import LeakyReLU\n",
+ "from keras.layers import Dropout\n",
+ "from keras.layers import Lambda\n",
+ "from keras.layers import Activation\n",
+ "from keras.layers import Concatenate\n",
+ "from keras.layers import BatchNormalization\n",
+ "from keras.metrics import SparseTopKCategoricalAccuracy\n",
+ "from keras.metrics import SparseCategoricalAccuracy\n",
+ "from keras.metrics import BinaryAccuracy\n",
+ "from keras.optimizers import Adam\n",
+ "from keras.applications import MobileNet\n",
+ "\n",
+ "import tensorflow as tf\n",
+ "import tensorflow_datasets.public_api as tfds\n",
+ "\n",
+ "import imgaug.augmenters as iaa\n",
+ "\n",
+ "from matplotlib import pyplot\n",
+ "\n",
+ "from rich.progress import track"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "IMAGE_SIZE = 64\n",
+ "LATENT_DIM = 512\n",
+ "BATCH_SIZE = 128\n",
+ "LEARNING_RATE = 3e-5\n",
+ "\n",
+ "PATH = '/content/projsemisup/'\n",
+ "CLASSES = os.listdir(PATH + 'Lab/')\n",
+ "NB_CLASSES = len(CLASSES)\n",
+ "LAB_COUNT = len(os.listdir(PATH + 'Lab/' + CLASSES[0] + '/'))\n",
+ "TEST_COUNT = len(os.listdir(PATH + 'Test/' + CLASSES[0] + '/'))"
+ ],
+ "metadata": {
+ "id": "86O4uJQYY8Vd"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def load_semisup_data():\n",
+ "\n",
+ " classes = os.listdir(PATH + 'Lab/')\n",
+ "\n",
+ " # Initialise les structures de données\n",
+ " x_lab = np.zeros((LAB_COUNT * NB_CLASSES, IMAGE_SIZE, IMAGE_SIZE, 3), dtype=np.uint8)\n",
+ " y_lab = np.zeros((LAB_COUNT * NB_CLASSES, 1))\n",
+ " i = 0\n",
+ " for c in track(classes, description='Labs...'):\n",
+ "\n",
+ " class_label = classes.index(c)\n",
+ " list_images = os.listdir(PATH + 'Lab/' + c + '/')\n",
+ "\n",
+ " for img_name in list_images:\n",
+ " # Lecture de l'image\n",
+ " img = open(PATH + 'Lab/' + c + '/' + img_name)\n",
+ "\n",
+ " # Mise à l'échelle de l'image\n",
+ " img = img.resize((IMAGE_SIZE, IMAGE_SIZE), ANTIALIAS)\n",
+ " img = img.convert('RGB')\n",
+ "\n",
+ " # Remplissage de la variable x\n",
+ " x_lab[i] = np.asarray(img, dtype=np.uint8)\n",
+ " y_lab[i] = class_label\n",
+ " i = i + 1\n",
+ "\n",
+ " list_images = os.listdir(PATH + 'Unlab/')\n",
+ " nb_unlab = len(list_images)\n",
+ " x_unlab = np.zeros((nb_unlab, IMAGE_SIZE, IMAGE_SIZE, 3), dtype=np.uint8)\n",
+ " i = 0\n",
+ " for img_name in track(list_images, description='Unlabs...'):\n",
+ " # Lecture de l'image\n",
+ " img = open(PATH + 'Unlab/' + img_name)\n",
+ "\n",
+ " # Mise à l'échelle de l'image\n",
+ " img = img.resize((IMAGE_SIZE, IMAGE_SIZE), ANTIALIAS)\n",
+ " img = img.convert('RGB')\n",
+ "\n",
+ " # Remplissage de la variable x\n",
+ " x_unlab[i] = np.asarray(img, dtype=np.uint8)\n",
+ " i = i + 1\n",
+ "\n",
+ " file_PATH_test = os.listdir(PATH + 'Test/')\n",
+ "\n",
+ " # Initialise les structures de données\n",
+ " x_test = np.zeros((TEST_COUNT * NB_CLASSES, IMAGE_SIZE, IMAGE_SIZE, 3), dtype=np.uint8)\n",
+ " y_test = np.zeros((TEST_COUNT * NB_CLASSES, 1))\n",
+ " i = 0\n",
+ " for c in track(file_PATH_test, description='Tests...'):\n",
+ "\n",
+ " class_label = classes.index(c)\n",
+ " list_images = os.listdir(PATH + 'Test/' + c + '/')\n",
+ "\n",
+ " for img_name in list_images:\n",
+ " # Lecture de l'image\n",
+ " img = open(PATH + 'Test/' + c + '/' + img_name, )\n",
+ "\n",
+ " # Mise à l'échelle de l'image\n",
+ " img = img.resize((IMAGE_SIZE, IMAGE_SIZE), ANTIALIAS)\n",
+ " img = img.convert('RGB')\n",
+ "\n",
+ " # Remplissage de la variable x\n",
+ " x_test[i] = np.asarray(img, dtype=np.uint8)\n",
+ " y_test[i] = class_label\n",
+ " i = i + 1\n",
+ "\n",
+ " return (\n",
+ " x_lab, y_lab,\n",
+ " x_unlab,\n",
+ " x_test, y_test\n",
+ " )"
+ ],
+ "metadata": {
+ "id": "UwadDSV1Y_Sc"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def define_discriminator():\n",
+ "\n",
+ " in_image = Input(shape=(IMAGE_SIZE, IMAGE_SIZE, 3))\n",
+ " fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(in_image)\n",
+ " fe = LeakyReLU(alpha=0.2)(fe)\n",
+ " fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(fe)\n",
+ " fe = LeakyReLU(alpha=0.2)(fe)\n",
+ " fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(fe)\n",
+ " fe = LeakyReLU(alpha=0.2)(fe)\n",
+ " fe = Conv2D(128, (3, 3), strides=(2, 2), padding='same')(fe)\n",
+ " fe = LeakyReLU(alpha=0.2)(fe)\n",
+ " fe = Flatten()(fe)\n",
+ " fe = Dropout(0.4)(fe)\n",
+ " fe = Dense(128)(fe)\n",
+ "\n",
+ " # mbn = MobileNet(\n",
+ " # input_shape=(IMAGE_SIZE, IMAGE_SIZE, 3),\n",
+ " # classes=NB_CLASSES,\n",
+ " # include_top=False,\n",
+ " # weights=None\n",
+ " # )\n",
+ " # fe = Flatten()(mbn.output)\n",
+ "\n",
+ " # Unsupervised output\n",
+ " d_out_layer = Dense(1)(fe)\n",
+ " d_out_layer = Activation('sigmoid')(d_out_layer)\n",
+ " d_model = Model(in_image, d_out_layer)\n",
+ " # d_model = Model(mbn.input, d_out_layer)\n",
+ " opt = Adam(learning_rate=LEARNING_RATE)\n",
+ " d_model.compile(loss='binary_crossentropy',\n",
+ " optimizer=opt, metrics=[BinaryAccuracy()])\n",
+ "\n",
+ " # Supervised output\n",
+ " c_out_layer = Dense(NB_CLASSES)(fe)\n",
+ " c_out_layer = Activation('softmax')(c_out_layer)\n",
+ " c_model = Model(in_image, c_out_layer)\n",
+ " # c_model = Model(mbn.input, c_out_layer)\n",
+ " opt = Adam(learning_rate=LEARNING_RATE)\n",
+ " c_model.compile(loss='sparse_categorical_crossentropy',\n",
+ " optimizer=opt, metrics=[SparseCategoricalAccuracy(), SparseTopKCategoricalAccuracy(k=3)])\n",
+ "\n",
+ " return d_model, c_model"
+ ],
+ "metadata": {
+ "id": "mIfaYQKnZE2T"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def define_generator():\n",
+ "\n",
+ " in_lat = Input(shape=(LATENT_DIM,))\n",
+ " gen = Dense(4*4*256)(in_lat)\n",
+ " gen = BatchNormalization(momentum=0.8)(gen)\n",
+ " gen = LeakyReLU(alpha=0.2)(gen)\n",
+ " gen = Reshape((4, 4, 256))(gen)\n",
+ " gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n",
+ " gen = BatchNormalization(momentum=0.8)(gen)\n",
+ " gen = LeakyReLU(alpha=0.2)(gen)\n",
+ " gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n",
+ " gen = BatchNormalization(momentum=0.8)(gen)\n",
+ " gen = LeakyReLU(alpha=0.2)(gen)\n",
+ " gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n",
+ " gen = BatchNormalization(momentum=0.8)(gen)\n",
+ " gen = LeakyReLU(alpha=0.2)(gen)\n",
+ " gen = Conv2DTranspose(128, 4, strides=2, padding=\"same\")(gen)\n",
+ " gen = BatchNormalization(momentum=0.8)(gen)\n",
+ " gen = LeakyReLU(alpha=0.2)(gen)\n",
+ " out_layer = Conv2D(3, (3, 3), padding=\"same\", activation=\"tanh\")(gen)\n",
+ "\n",
+ " model = Model(in_lat, out_layer)\n",
+ "\n",
+ " return model"
+ ],
+ "metadata": {
+ "id": "5Bcd3eRCZHkG"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def define_gan(g_model, d_model):\n",
+ "\n",
+ " d_model.trainable = False\n",
+ " gan_output = d_model(g_model.output)\n",
+ "\n",
+ " model = Model(g_model.input, gan_output)\n",
+ "\n",
+ " opt = Adam(learning_rate=2*LEARNING_RATE)\n",
+ " model.compile(loss='binary_crossentropy',\n",
+ " optimizer=opt, metrics=[BinaryAccuracy()])\n",
+ "\n",
+ " return model"
+ ],
+ "metadata": {
+ "id": "0cAFqWRCZJLW"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def select_supervised_samples(dataset, n_samples=100):\n",
+ " X, y = dataset\n",
+ " X_list, y_list = list(), list()\n",
+ " n_per_class = int(n_samples / NB_CLASSES)\n",
+ " for i in range(NB_CLASSES):\n",
+ " # get all images for this class\n",
+ " X_with_class = X[y == i]\n",
+ " # choose random instances\n",
+ " ix = randint(0, len(X_with_class), n_per_class)\n",
+ " # add to list\n",
+ " [X_list.append(X_with_class[j]) for j in ix]\n",
+ " [y_list.append(i) for j in ix]\n",
+ " return asarray(X_list), asarray(y_list)"
+ ],
+ "metadata": {
+ "id": "pSDIItbJZLaR"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def generate_real_samples(dataset, n_samples, aug=None):\n",
+ " # split into images and labels\n",
+ " images, labels = dataset\n",
+ " # choose random instances\n",
+ " ix = randint(0, images.shape[0], n_samples)\n",
+ " # select images and labels\n",
+ " X = images[ix]\n",
+ " if labels is not None:\n",
+ " labels = labels[ix]\n",
+ " # generate class labels\n",
+ " y = ones((n_samples, 1))\n",
+ " # apply augmentation\n",
+ " if aug is not None:\n",
+ " X = aug(images=X)\n",
+ " # normalize batch\n",
+ " X = X.astype(\"double\") / 127.5 - 1.0\n",
+ " return X, labels, y"
+ ],
+ "metadata": {
+ "id": "x_ZiI1KdZNVT"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def generate_latent_points(n_samples):\n",
+ " # generate points in the latent space\n",
+ " z_input = randn(n_samples * LATENT_DIM)\n",
+ " # reshape into a batch of inputs for the network\n",
+ " z_input = z_input.reshape(n_samples, LATENT_DIM)\n",
+ " return z_input"
+ ],
+ "metadata": {
+ "id": "lkUFZ6xOZPJS"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def generate_fake_samples(generator, n_samples):\n",
+ " # generate points in latent space\n",
+ " z_input = generate_latent_points(n_samples)\n",
+ " # predict outputs\n",
+ " images = generator.predict(z_input, verbose=0)\n",
+ " # create class labels\n",
+ " y = zeros((n_samples, 1))\n",
+ " return images, y"
+ ],
+ "metadata": {
+ "id": "XvOWaVWsZQla"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def summarize_performance(step, d_model, g_model, c_model, dataset, n_samples=100):\n",
+ " # prepare fake examples\n",
+ " X, _ = generate_fake_samples(g_model, n_samples)\n",
+ " # scale from [-1,1] to [0,1]\n",
+ " X = (X + 1) / 2.0\n",
+ " # create figure\n",
+ " pyplot.figure(figsize=(10, 4))\n",
+ " # plot images\n",
+ " for i in range(10):\n",
+ " # define subplot\n",
+ " pyplot.subplot(2, 5, 1 + i)\n",
+ " # turn off axis\n",
+ " pyplot.axis('off')\n",
+ " # plot raw pixel data\n",
+ " pyplot.imshow(X[i, :, :, :])\n",
+ " pyplot.show()\n",
+ " \n",
+ " # get test images\n",
+ " _, _, _, x_test, y_test = dataset\n",
+ " # normalize images\n",
+ " x_test = x_test.astype(\"double\") / 127.5 - 1.0\n",
+ " # evaluate the classifier model\n",
+ " _, acc, acc3 = c_model.evaluate(x_test, y_test, verbose=0)\n",
+ " print(f\"Test accuracy:\\n top 1: {acc*100:.3f}%\\n top 3: {acc3*100:.3f}%\")\n",
+ " # save the discriminator model\n",
+ " filename1 = f\"d_model_{step+1:04d}.h5\"\n",
+ " d_model.save_weights(filename1)\n",
+ " # save the generator model\n",
+ " filename2 = f\"g_model_{step+1:04d}.h5\"\n",
+ " g_model.save_weights(filename2)\n",
+ " # save the classifier model\n",
+ " filename3 = f\"c_model_{step+1:04d}.h5\"\n",
+ " c_model.save_weights(filename3)\n",
+ " print(f\"Saved: {filename1}, {filename2}, {filename3}\\n\")\n",
+ "\n",
+ " return acc, acc3"
+ ],
+ "metadata": {
+ "id": "6xybda-3ZTFc"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def train(g_model, d_model, c_model, gan_model, dataset, aug=None, nb_epochs=100):\n",
+ " # select supervised dataset\n",
+ " x_lab, y_lab, x_unlab, _, _ = dataset\n",
+ " # calculate the number of batches per training epoch\n",
+ " bat_per_epo = int(x_unlab.shape[0] / BATCH_SIZE)\n",
+ " # store accuracies\n",
+ " accuracies = {'train': {'top1': [], 'top3': []}, 'val': {'top1': [], 'top3': []}}\n",
+ "\n",
+ " print(f\"nb_epochs={nb_epochs}, batch_size={BATCH_SIZE}, b/e={bat_per_epo}\\n\")\n",
+ " for epoch in range(nb_epochs):\n",
+ "\n",
+ " print(f\"epoch {epoch+1:3}/{nb_epochs}:\")\n",
+ " for step in range(bat_per_epo):\n",
+ " # update supervised discriminator (c)\n",
+ " X, Y, _ = generate_real_samples([x_lab, y_lab], BATCH_SIZE, aug)\n",
+ " c_loss, c_acc, c_acc3 = c_model.train_on_batch(X, Y)\n",
+ " # update unsupervised discriminator (d)\n",
+ " X_real, _, Y_real = generate_real_samples([x_unlab, None], BATCH_SIZE//2)\n",
+ " X_fake, Y_fake = generate_fake_samples(g_model, BATCH_SIZE//2)\n",
+ " X = np.concatenate([X_real, X_fake], axis=0)\n",
+ " Y = np.concatenate([Y_real, Y_fake], axis=0)\n",
+ " d_loss, d_acc = d_model.train_on_batch(X, Y)\n",
+ " # update generator (g)\n",
+ " X_gan, y_gan = generate_latent_points(BATCH_SIZE), ones((BATCH_SIZE, 1))\n",
+ " g_loss, g_acc = gan_model.train_on_batch(X_gan, y_gan)\n",
+ " # show losses and accuracies\n",
+ " print(f\"\\rstep {step+1:5}/{bat_per_epo}:\" +\n",
+ " f\" c[{c_loss:7.3f}, {c_acc*100:3.0f}%, {c_acc3*100:3.0f}%]\" +\n",
+ " f\" d[{d_loss:7.3f}, {d_acc*100:3.0f}%]\" +\n",
+ " f\" g[{g_loss:7.3f}, {g_acc*100:3.0f}%]\",\n",
+ " end='')\n",
+ " # evaluate the model\n",
+ " acc, acc3 = summarize_performance(epoch, d_model, g_model, c_model, dataset)\n",
+ " accuracies['train']['top1'].append(c_acc)\n",
+ " accuracies['train']['top3'].append(c_acc3)\n",
+ " accuracies['val']['top1'].append(acc)\n",
+ " accuracies['val']['top3'].append(acc3)\n",
+ " print(accuracies)"
+ ],
+ "metadata": {
+ "id": "2MoC8sysY0Lc"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# load image data\n",
+ "dataset = load_semisup_data()"
+ ],
+ "metadata": {
+ "id": "_GU7lrzCaVzF",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 65,
+ "referenced_widgets": [
+ "1b3a5e8e274742b5a04af83869731bb0",
+ "91c8bbbabb28486f9e14f5ddda207e52",
+ "f4b81afd77a64eff9395edc9181c531a",
+ "75dea27498a14b6d85cba75e3661214d",
+ "c8b4fb345aa44aa7b3b5b373cab3ba9c",
+ "93229082acb045e980e0c34a7a60ba6a"
+ ]
+ },
+ "outputId": "ff2a09af-7780-47f1-9e26-e47b5e39b591"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [
+ "Output()"
+ ],
+ "application/vnd.jupyter.widget-view+json": {
+ "version_major": 2,
+ "version_minor": 0,
+ "model_id": "1b3a5e8e274742b5a04af83869731bb0"
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [],
+ "text/html": [
+ "\n"
+ ]
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [
+ "\n"
+ ],
+ "text/html": [
+ "\n",
+ "
\n"
+ ]
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [
+ "Output()"
+ ],
+ "application/vnd.jupyter.widget-view+json": {
+ "version_major": 2,
+ "version_minor": 0,
+ "model_id": "f4b81afd77a64eff9395edc9181c531a"
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [],
+ "text/html": [
+ "\n"
+ ]
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [
+ "\n"
+ ],
+ "text/html": [
+ "\n",
+ "
\n"
+ ]
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [
+ "Output()"
+ ],
+ "application/vnd.jupyter.widget-view+json": {
+ "version_major": 2,
+ "version_minor": 0,
+ "model_id": "c8b4fb345aa44aa7b3b5b373cab3ba9c"
+ }
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [],
+ "text/html": [
+ "\n"
+ ]
+ },
+ "metadata": {}
+ },
+ {
+ "output_type": "display_data",
+ "data": {
+ "text/plain": [
+ "\n"
+ ],
+ "text/html": [
+ "\n",
+ "
\n"
+ ]
+ },
+ "metadata": {}
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# augmentation\n",
+ "aug = iaa.RandAugment(n=2, m=9)"
+ ],
+ "metadata": {
+ "id": "vlC87HYvWUY8"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# create the discriminator models\n",
+ "d_model, c_model = define_discriminator()\n",
+ "d_model.summary()\n",
+ "c_model.summary()\n",
+ "# create the generator\n",
+ "g_model = define_generator()\n",
+ "g_model.summary()\n",
+ "# create the gan\n",
+ "gan_model = define_gan(g_model, d_model)\n",
+ "# train model\n",
+ "train(g_model, d_model, c_model, gan_model, dataset, aug)"
+ ],
+ "metadata": {
+ "id": "zuWf9tXqZVE-",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 1000
+ },
+ "outputId": "a28e2ee4-5a71-4340-bfaf-c1d21457740b"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "Model: \"model_4\"\n",
+ "_________________________________________________________________\n",
+ " Layer (type) Output Shape Param # \n",
+ "=================================================================\n",
+ " input_3 (InputLayer) [(None, 64, 64, 3)] 0 \n",
+ " \n",
+ " conv2d_8 (Conv2D) (None, 32, 32, 128) 3584 \n",
+ " \n",
+ " leaky_re_lu_8 (LeakyReLU) (None, 32, 32, 128) 0 \n",
+ " \n",
+ " conv2d_9 (Conv2D) (None, 16, 16, 128) 147584 \n",
+ " \n",
+ " leaky_re_lu_9 (LeakyReLU) (None, 16, 16, 128) 0 \n",
+ " \n",
+ " conv2d_10 (Conv2D) (None, 8, 8, 128) 147584 \n",
+ " \n",
+ " leaky_re_lu_10 (LeakyReLU) (None, 8, 8, 128) 0 \n",
+ " \n",
+ " conv2d_11 (Conv2D) (None, 4, 4, 128) 147584 \n",
+ " \n",
+ " leaky_re_lu_11 (LeakyReLU) (None, 4, 4, 128) 0 \n",
+ " \n",
+ " flatten_2 (Flatten) (None, 2048) 0 \n",
+ " \n",
+ " dropout_2 (Dropout) (None, 2048) 0 \n",
+ " \n",
+ " dense_6 (Dense) (None, 128) 262272 \n",
+ " \n",
+ " dense_7 (Dense) (None, 1) 129 \n",
+ " \n",
+ " activation_4 (Activation) (None, 1) 0 \n",
+ " \n",
+ "=================================================================\n",
+ "Total params: 708,737\n",
+ "Trainable params: 708,737\n",
+ "Non-trainable params: 0\n",
+ "_________________________________________________________________\n",
+ "Model: \"model_5\"\n",
+ "_________________________________________________________________\n",
+ " Layer (type) Output Shape Param # \n",
+ "=================================================================\n",
+ " input_3 (InputLayer) [(None, 64, 64, 3)] 0 \n",
+ " \n",
+ " conv2d_8 (Conv2D) (None, 32, 32, 128) 3584 \n",
+ " \n",
+ " leaky_re_lu_8 (LeakyReLU) (None, 32, 32, 128) 0 \n",
+ " \n",
+ " conv2d_9 (Conv2D) (None, 16, 16, 128) 147584 \n",
+ " \n",
+ " leaky_re_lu_9 (LeakyReLU) (None, 16, 16, 128) 0 \n",
+ " \n",
+ " conv2d_10 (Conv2D) (None, 8, 8, 128) 147584 \n",
+ " \n",
+ " leaky_re_lu_10 (LeakyReLU) (None, 8, 8, 128) 0 \n",
+ " \n",
+ " conv2d_11 (Conv2D) (None, 4, 4, 128) 147584 \n",
+ " \n",
+ " leaky_re_lu_11 (LeakyReLU) (None, 4, 4, 128) 0 \n",
+ " \n",
+ " flatten_2 (Flatten) (None, 2048) 0 \n",
+ " \n",
+ " dropout_2 (Dropout) (None, 2048) 0 \n",
+ " \n",
+ " dense_6 (Dense) (None, 128) 262272 \n",
+ " \n",
+ " dense_8 (Dense) (None, 18) 2322 \n",
+ " \n",
+ " activation_5 (Activation) (None, 18) 0 \n",
+ " \n",
+ "=================================================================\n",
+ "Total params: 710,930\n",
+ "Trainable params: 710,930\n",
+ "Non-trainable params: 0\n",
+ "_________________________________________________________________\n",
+ "Model: \"model_6\"\n",
+ "_________________________________________________________________\n",
+ " Layer (type) Output Shape Param # \n",
+ "=================================================================\n",
+ " input_4 (InputLayer) [(None, 512)] 0 \n",
+ " \n",
+ " dense_9 (Dense) (None, 4096) 2101248 \n",
+ " \n",
+ " batch_normalization (BatchN (None, 4096) 16384 \n",
+ " ormalization) \n",
+ " \n",
+ " leaky_re_lu_12 (LeakyReLU) (None, 4096) 0 \n",
+ " \n",
+ " reshape (Reshape) (None, 4, 4, 256) 0 \n",
+ " \n",
+ " conv2d_transpose (Conv2DTra (None, 8, 8, 128) 524416 \n",
+ " nspose) \n",
+ " \n",
+ " batch_normalization_1 (Batc (None, 8, 8, 128) 512 \n",
+ " hNormalization) \n",
+ " \n",
+ " leaky_re_lu_13 (LeakyReLU) (None, 8, 8, 128) 0 \n",
+ " \n",
+ " conv2d_transpose_1 (Conv2DT (None, 16, 16, 128) 262272 \n",
+ " ranspose) \n",
+ " \n",
+ " batch_normalization_2 (Batc (None, 16, 16, 128) 512 \n",
+ " hNormalization) \n",
+ " \n",
+ " leaky_re_lu_14 (LeakyReLU) (None, 16, 16, 128) 0 \n",
+ " \n",
+ " conv2d_transpose_2 (Conv2DT (None, 32, 32, 128) 262272 \n",
+ " ranspose) \n",
+ " \n",
+ " batch_normalization_3 (Batc (None, 32, 32, 128) 512 \n",
+ " hNormalization) \n",
+ " \n",
+ " leaky_re_lu_15 (LeakyReLU) (None, 32, 32, 128) 0 \n",
+ " \n",
+ " conv2d_transpose_3 (Conv2DT (None, 64, 64, 128) 262272 \n",
+ " ranspose) \n",
+ " \n",
+ " batch_normalization_4 (Batc (None, 64, 64, 128) 512 \n",
+ " hNormalization) \n",
+ " \n",
+ " leaky_re_lu_16 (LeakyReLU) (None, 64, 64, 128) 0 \n",
+ " \n",
+ " conv2d_12 (Conv2D) (None, 64, 64, 3) 3459 \n",
+ " \n",
+ "=================================================================\n",
+ "Total params: 3,434,371\n",
+ "Trainable params: 3,425,155\n",
+ "Non-trainable params: 9,216\n",
+ "_________________________________________________________________\n"
+ ]
+ },
+ {
+ "output_type": "error",
+ "ename": "NameError",
+ "evalue": "ignored",
+ "traceback": [
+ "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
+ "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
+ "\u001b[0;32m\u001b[0m in \u001b[0;36m\u001b[0;34m\u001b[0m\n\u001b[1;32m 9\u001b[0m \u001b[0mgan_model\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdefine_gan\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_model\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0md_model\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 10\u001b[0m \u001b[0;31m# train model\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 11\u001b[0;31m \u001b[0mtrain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mg_model\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0md_model\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mc_model\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgan_model\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdataset\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maug\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
+ "\u001b[0;31mNameError\u001b[0m: name 'train' is not defined"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "backup_name = '0040'\n",
+ "\n",
+ "# create the discriminator models\n",
+ "d_model, c_model = define_discriminator()\n",
+ "# create the generator\n",
+ "g_model = define_generator()\n",
+ "\n",
+ "d_model.load_weights(f\"d_model_{backup_name}.h5\")\n",
+ "c_model.load_weights(f\"c_model_{backup_name}.h5\")\n",
+ "g_model.load_weights(f\"g_model_{backup_name}.h5\")\n",
+ "\n",
+ "# create the gan\n",
+ "gan_model = define_gan(g_model, d_model)\n",
+ "\n",
+ "# create the gan\n",
+ "gan_model = define_gan(g_model, d_model)\n",
+ "# train model\n",
+ "train(g_model, d_model, c_model, gan_model, dataset, aug)"
+ ],
+ "metadata": {
+ "id": "7btFJZgciMoR",
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 1000
+ },
+ "outputId": "17cc1f09-88dc-4080-ec60-fb07b874e32d"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "metadata": {
+ "tags": null
+ },
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "nb_epochs=100, batch_size=128, b/e=161\n",
+ "\n",
+ "epoch 1/100:\n",
+ "step 161/161: c[ 0.096, 97%, 99%] d[ 0.291, 88%] g[ 2.927, 6%]"
+ ]
+ },
+ {
+ "data": {
+ "image/png": "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\n",
+ "text/plain": [
+ "