{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# ImageNet Classifier with VGG16\nDon't forget to upvote 💚\n* [I. Load Data](#load_data)\n    - [I.1. Image transformation during preprocessing](#show_data_preprocessing)\n* [II. Import the pre-trained VGG16 model](#import_VGG16)\n    - [II.1. Architecture of the VGG16](#VGG16_architecture)\n    - [II.2. Load model weights](#load_weights_model)\n* [III. Predictions on the training set](#predictions_on_training_set)\n    - [III.1. Predictions](#predictions)\n    - [III.2. Accuracy](#accuracy)\n    - [III.3. Show images with predicted and actual class](#show_data_prediction)\n* [IV. Transfer learning / Monkey classification](#transfer_learning)\n    - [IV.1. Load data](#load_data_1)\n    - [IV.2. Create monkey_model from VGG16](#monkey_model)\n    - [IV.3. Train the model](#training)\n    - [IV.4. Predictions on test_set](#predictions_test_set)","metadata":{}},{"cell_type":"code","source":"#Install dependecies\nimport numpy as np\nimport pandas as pd\nfrom tqdm import tqdm\nimport os\nfrom matplotlib import pyplot as plt\nimport random\nfrom glob import glob\n\nimport tensorflow as tf\nimport keras\nfrom keras.models import *\nfrom keras.layers import *\nfrom tensorflow.keras.callbacks import EarlyStopping\nfrom tensorflow.keras.optimizers import Adam\nfrom keras.applications.vgg16 import preprocess_input\nfrom keras.preprocessing.image import ImageDataGenerator, load_img, img_to_array, array_to_img\nfrom keras.applications.vgg16 import VGG16\nfrom IPython.display import display\nfrom PIL import Image","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:45:37.47222Z","iopub.execute_input":"2023-02-01T14:45:37.472697Z","iopub.status.idle":"2023-02-01T14:45:37.486123Z","shell.execute_reply.started":"2023-02-01T14:45:37.472658Z","shell.execute_reply":"2023-02-01T14:45:37.484932Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"load_data\"></a>\n# I. Load Data\n* train_data : 1000 classes de 1300 datas each\n* test_data : 100 000 datas","metadata":{"execution":{"iopub.status.busy":"2023-01-27T20:52:17.868052Z","iopub.execute_input":"2023-01-27T20:52:17.868617Z","iopub.status.idle":"2023-01-27T20:52:17.876642Z","shell.execute_reply.started":"2023-01-27T20:52:17.868573Z","shell.execute_reply":"2023-01-27T20:52:17.875286Z"}}},{"cell_type":"code","source":"mapping_path = '/kaggle/input/imagenet-object-localization-challenge/LOC_synset_mapping.txt'\n\n# Creation of mapping dictionaries to obtain the image classes\n\nclass_mapping_dict = {}\nclass_mapping_dict_number = {}\nmapping_class_to_number = {}\nmapping_number_to_class = {}\ni = 0\nfor line in open(mapping_path):\n    class_mapping_dict[line[:9].strip()] = line[9:].strip()\n    class_mapping_dict_number[i] = line[9:].strip()\n    mapping_class_to_number[line[:9].strip()] = i\n    mapping_number_to_class[i] = line[:9].strip()\n    i+=1","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:45:39.410818Z","iopub.execute_input":"2023-02-01T14:45:39.4113Z","iopub.status.idle":"2023-02-01T14:45:39.444474Z","shell.execute_reply.started":"2023-02-01T14:45:39.411219Z","shell.execute_reply":"2023-02-01T14:45:39.443539Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_path = '/kaggle/input/imagenet-object-localization-challenge/ILSVRC/Data/CLS-LOC/train'\n\n# Creation of dataset_array and true_classes\n\ndataset_array = []\ntrue_classes = []\nimages_array = []\nfor train_class in tqdm(os.listdir(train_path)):\n    i = 0\n    for el in os.listdir(train_path + '/' + train_class):\n        if i < 10:\n            path = train_path + '/' + train_class + '/' + el\n            image = load_img(path,target_size=(224,224,3))\n            image_array = img_to_array(image).astype(np.uint8)\n            images_array.append(image_array)\n            true_class = class_mapping_dict[path.split('/')[-2]]\n            true_classes.append(true_class)\n            i+=1\n        else:\n            break\nimages_array = np.array(images_array)\ntrue_classes = np.array(true_classes)\nprint('Preprocessing in progress')\ndataset_array = preprocess_input(images_array)\nprint('FINISH')","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:47:49.817767Z","iopub.execute_input":"2023-02-01T14:47:49.818131Z","iopub.status.idle":"2023-02-01T14:57:53.603864Z","shell.execute_reply.started":"2023-02-01T14:47:49.818099Z","shell.execute_reply":"2023-02-01T14:57:53.602919Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"show_data_preprocessing\"></a>\n<font size=\"3.5\">I.1. Image transformation during preprocessing</font>","metadata":{}},{"cell_type":"code","source":"random_index = random.sample(range(0, 1000), 5)\n\nplt.figure(figsize=(20, 20))\nplt.suptitle('Before preprocessing', x = 0.5, y = 0.6)\nfor i in range(5):\n    ax = plt.subplot(1, 5, i + 1)\n    ax.imshow(images_array[random_index[i]])\n    \nplt.figure(figsize=(20, 20))\nplt.suptitle('After preprocessing', x = 0.5, y = 0.6)\nfor i in range(5):\n    ax = plt.subplot(1, 5, i + 1)\n    ax.imshow(dataset_array[random_index[i]])","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:58:24.891692Z","iopub.execute_input":"2023-02-01T14:58:24.892099Z","iopub.status.idle":"2023-02-01T14:58:26.470082Z","shell.execute_reply.started":"2023-02-01T14:58:24.892064Z","shell.execute_reply":"2023-02-01T14:58:26.469155Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"import_VGG16\"></a>\n# II. Import the pre-trained VGG16 model","metadata":{}},{"cell_type":"markdown","source":"<a id=\"VGG16_architecture\"></a>\n<font size=\"3.5\">II.1. Architecture of the VGG16</font>","metadata":{}},{"cell_type":"markdown","source":"![Capture d’écran 2023-01-28 001326.png](attachment:ee78af2a-900c-462e-a0ac-cb78450e0b90.png)","metadata":{"execution":{"iopub.status.busy":"2023-01-27T23:10:05.585867Z","iopub.execute_input":"2023-01-27T23:10:05.586397Z","iopub.status.idle":"2023-01-27T23:10:05.593699Z","shell.execute_reply.started":"2023-01-27T23:10:05.586358Z","shell.execute_reply":"2023-01-27T23:10:05.592124Z"}},"attachments":{"ee78af2a-900c-462e-a0ac-cb78450e0b90.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"<a id=\"load_weights_model\"></a>\n<font size=\"3.5\">II.2. Load model weights</font>","metadata":{}},{"cell_type":"code","source":"model_VGG16 = VGG16(weights='imagenet', include_top=True, input_shape=(224,224,3))\nmodel_VGG16.summary()","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:58:51.808073Z","iopub.execute_input":"2023-02-01T14:58:51.808448Z","iopub.status.idle":"2023-02-01T14:58:53.634556Z","shell.execute_reply.started":"2023-02-01T14:58:51.808415Z","shell.execute_reply":"2023-02-01T14:58:53.633528Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"predictions_on_training_set\"></a>\n# III. Predictions on the training set","metadata":{}},{"cell_type":"markdown","source":"<a id=\"predictions\"></a>\n<font size=\"3.5\">III.1. Predictions</font>","metadata":{}},{"cell_type":"code","source":"n = 1000\n\nprint('Prediction in progress')\npredictions_array = model_VGG16.predict(dataset_array[:n])\nprint('Prediction OK')\n\npredict_classes = []\n\nfor i in tqdm(range(n)):\n    arg_max = predictions_array[i].argmax()\n    predict_class = class_mapping_dict_number[arg_max]\n    predict_classes.append(predict_class)\n    \npredict_classes = np.array(predict_classes)","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:59:02.200078Z","iopub.execute_input":"2023-02-01T14:59:02.200603Z","iopub.status.idle":"2023-02-01T14:59:25.446927Z","shell.execute_reply.started":"2023-02-01T14:59:02.200561Z","shell.execute_reply":"2023-02-01T14:59:25.445957Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"accuracy\"></a>\n<font size=\"3.5\">III.2. Accuracy</font>","metadata":{}},{"cell_type":"code","source":"accuracy = 0\nfor i in range(n):\n    if predict_classes[i] == true_classes[i]:\n        accuracy+=1\naccuracy /= n\nprint('accuracy : ' + str(accuracy))","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:59:49.46689Z","iopub.execute_input":"2023-02-01T14:59:49.467267Z","iopub.status.idle":"2023-02-01T14:59:49.47523Z","shell.execute_reply.started":"2023-02-01T14:59:49.46722Z","shell.execute_reply":"2023-02-01T14:59:49.474264Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"show_data_prediction\"></a>\n<font size=\"3.5\">III.3. Show images with predicted and actual class</font>","metadata":{}},{"cell_type":"code","source":"random_index = random.sample(range(0, n), 5)\n\nplt.figure(figsize=(25, 25))\nplt.suptitle('Predict classes', x = 0.5, y = 0.6)\nfor i in range(5):\n    ax = plt.subplot(1, 5, i + 1)\n    ax.imshow(images_array[random_index[i]])\n    if predict_classes[random_index[i]] == true_classes[random_index[i]]:\n        plt.title(predict_classes[random_index[i]], color='green', size=10)\n    else:\n        plt.title(predict_classes[random_index[i]], color='red', size=10)\n\nplt.figure(figsize=(25, 25))\nplt.suptitle('True classes', x = 0.5, y = 0.6)\nfor i in range(5):\n    ax = plt.subplot(1, 5, i + 1)\n    ax.imshow(images_array[random_index[i]])\n    if predict_classes[random_index[i]] == true_classes[random_index[i]]:\n        plt.title(true_classes[random_index[i]], color='green', size=10)\n    else:\n        plt.title(true_classes[random_index[i]], color='red', size=10)","metadata":{"execution":{"iopub.status.busy":"2023-02-01T14:59:54.78884Z","iopub.execute_input":"2023-02-01T14:59:54.789201Z","iopub.status.idle":"2023-02-01T14:59:56.231118Z","shell.execute_reply.started":"2023-02-01T14:59:54.789172Z","shell.execute_reply":"2023-02-01T14:59:56.230291Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"transfer_learning\"></a>\n# IV. Transfer learning / Monkey classification\nWe keep the model and the weights of the layers before the Flatten layer. We just change the Dense layers, to adapt it to a monkey classification problem. We have 10 fruit classes so we need a vector of size 10 in the output.","metadata":{}},{"cell_type":"code","source":"#Install dependecies\nimport numpy as np\nimport pandas as pd\nfrom tqdm import tqdm\nimport os\nfrom matplotlib import pyplot as plt\nimport random\nimport cv2\nfrom glob import glob\n\nimport tensorflow as tf\nimport keras\nfrom keras.models import *\nfrom keras.layers import *\nfrom tensorflow.keras.callbacks import EarlyStopping\nfrom tensorflow.keras.optimizers import Adam\nfrom keras.applications.vgg16 import preprocess_input\nfrom keras.preprocessing.image import ImageDataGenerator, load_img, img_to_array, array_to_img\nfrom keras.applications.vgg16 import VGG16\nfrom IPython.display import display\nfrom PIL import Image","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:00:04.720381Z","iopub.execute_input":"2023-02-01T15:00:04.720744Z","iopub.status.idle":"2023-02-01T15:00:04.915742Z","shell.execute_reply.started":"2023-02-01T15:00:04.720713Z","shell.execute_reply":"2023-02-01T15:00:04.913327Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"load_data_1\"></a>\n<font size=\"5\">IV.1. Load data</font>","metadata":{}},{"cell_type":"code","source":"# paths and constants\n\nmapping_path = '/kaggle/input/10-monkey-species/monkey_labels.txt'\nsrc_path_train = \"/kaggle/input/10-monkey-species/training/training\"\nsrc_path_test = \"/kaggle/input/10-monkey-species/validation/validation\"\n\nbatch_size = 32\nepochs = 10","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:02:57.307521Z","iopub.execute_input":"2023-02-01T15:02:57.307907Z","iopub.status.idle":"2023-02-01T15:02:57.31343Z","shell.execute_reply.started":"2023-02-01T15:02:57.307875Z","shell.execute_reply":"2023-02-01T15:02:57.312439Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Mapping dico between the output vector and the 10 classes\n\nclass_mapping_dict = {}\ni = 0\nfor line in open(mapping_path):\n    if 1<=i<=10:\n        class_mapping_dict[i-1] = line.split(' , ')[2].strip()\n    i+=1","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:02:57.748824Z","iopub.execute_input":"2023-02-01T15:02:57.749761Z","iopub.status.idle":"2023-02-01T15:02:57.77066Z","shell.execute_reply.started":"2023-02-01T15:02:57.749706Z","shell.execute_reply":"2023-02-01T15:02:57.769754Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Creation of the train_generator and the test_generator\n\nimage_gen = ImageDataGenerator(\n    rescale=1 / 255.0,\n    rotation_range=20,\n    zoom_range=0.05,\n    width_shift_range=0.05,\n    height_shift_range=0.05,\n    shear_range=0.05,\n    horizontal_flip=True,\n    fill_mode=\"nearest\",\n    preprocessing_function = preprocess_input,\n    validation_split=0.20)\n\ntrain_generator = image_gen.flow_from_directory(\n  src_path_train,\n  target_size=(224,224),\n  shuffle=True,\n  batch_size=batch_size,\n)\n\ntest_generator = image_gen.flow_from_directory(\n  src_path_test,\n  target_size=(224,224),\n  shuffle=True,\n  batch_size=batch_size,\n)\n\n# Lists of the filepaths\n\ntrain_image_files = glob(src_path_train + '/*/*.jp*g')\ntest_image_files = glob(src_path_test + '/*/*.jp*g')","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:03:03.238189Z","iopub.execute_input":"2023-02-01T15:03:03.238808Z","iopub.status.idle":"2023-02-01T15:03:03.809627Z","shell.execute_reply.started":"2023-02-01T15:03:03.238774Z","shell.execute_reply":"2023-02-01T15:03:03.808669Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"monkey_model\"></a>\n<font size=\"5\">IV.2. Create monkey_model from VGG16</font>","metadata":{}},{"cell_type":"code","source":"model_VGG16 = VGG16(weights='imagenet', include_top=False, input_shape=(224,224,3))\n\nfor layer in model_VGG16.layers:\n        layer.trainable = False\n\nmonkey_model = Sequential()\nmonkey_model.add(model_VGG16)\n\n# We change the last layers of the original VGG16 to adapt it to our animal classification between 10 classes\nmonkey_model.add(Flatten(name='flatten'))\nmonkey_model.add(Dense(units=4096, activation='relu', name='fc1'))\nmonkey_model.add(Dense(units=4096, activation='relu', name='fc2'))\nmonkey_model.add(Dense(units=10, activation='softmax', name='output'))\n\nmonkey_model.compile(optimizer=Adam(learning_rate=0.001), loss=keras.losses.categorical_crossentropy, metrics=['accuracy'])\n\nmonkey_model.summary()","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:03:06.694054Z","iopub.execute_input":"2023-02-01T15:03:06.694435Z","iopub.status.idle":"2023-02-01T15:03:08.388603Z","shell.execute_reply.started":"2023-02-01T15:03:06.694402Z","shell.execute_reply":"2023-02-01T15:03:08.387411Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"training\"></a>\n<font size=\"5\">IV.3. Train the model</font>","metadata":{}},{"cell_type":"code","source":"early_stop = EarlyStopping(monitor='val_loss',patience=2)\n\nhistory = monkey_model.fit(\n  train_generator,\n  validation_data=test_generator,\n  epochs=epochs,\n  steps_per_epoch=len(train_image_files) // batch_size,\n  validation_steps=len(test_image_files) // batch_size,\n  callbacks=[early_stop]\n)","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:03:32.896964Z","iopub.execute_input":"2023-02-01T15:03:32.897355Z","iopub.status.idle":"2023-02-01T15:10:21.984652Z","shell.execute_reply.started":"2023-02-01T15:03:32.897322Z","shell.execute_reply":"2023-02-01T15:10:21.983344Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"metrics = ['loss','accuracy','val_loss','val_accuracy']\nplt.figure(figsize=(20, 4))\nfor i in range(len(metrics)):\n    ax = plt.subplot(1, len(metrics), i + 1)\n    ax.plot(history.history[metrics[i]])\n    plt.title(metrics[i])","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:16:31.316614Z","iopub.execute_input":"2023-02-01T15:16:31.317056Z","iopub.status.idle":"2023-02-01T15:16:31.760965Z","shell.execute_reply.started":"2023-02-01T15:16:31.317017Z","shell.execute_reply":"2023-02-01T15:16:31.760036Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"<a id=\"predictions_test_set\"></a>\n<font size=\"5\">IV.4. Predictions on test_set</font>","metadata":{}},{"cell_type":"code","source":"# Calcul de l'accuracy sur notre test_set\nevaluation_dico = monkey_model.evaluate(test_generator, batch_size = 32, verbose = 0, return_dict=True)\nprint('Monkey_model :')\nprint('Loss : ' + str(evaluation_dico['loss']))\nprint('Accuracy : ' + str(evaluation_dico['accuracy']))","metadata":{"execution":{"iopub.status.busy":"2023-02-01T16:15:03.550573Z","iopub.execute_input":"2023-02-01T16:15:03.550943Z","iopub.status.idle":"2023-02-01T16:15:14.832934Z","shell.execute_reply.started":"2023-02-01T16:15:03.550912Z","shell.execute_reply":"2023-02-01T16:15:14.831933Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# load a batch fromthe test_set\nfinal_test = test_generator.next()\nfinal_images = final_test[0]\nfinal_labels = final_test[1]\n\n# Predict the classes with monkey_model\npredict_classes = monkey_model.predict(final_images)\n\n# Plot the results\n\nplt.figure(figsize=(20, 20))\nplt.suptitle('Predict classes', x = 0.5, y = 0.65)\nfor i in range(5):\n    ax = plt.subplot(1, 5, i + 1)\n    ax.imshow(cv2.cvtColor(final_images[i], cv2.COLOR_BGR2RGB))\n    plt.axis('off')\n    if predict_classes[i].argmax() == final_labels[i].argmax() :\n        plt.title(class_mapping_dict[predict_classes[i].argmax()], c='g')\n    else:\n        plt.title(class_mapping_dict[predict_classes[i].argmax()], c='r')\n\nplt.figure(figsize=(20, 20))\nplt.suptitle('True classes', x = 0.5, y = 0.65)\nfor i in range(5):\n    ax = plt.subplot(1, 5, i + 1)\n    ax.imshow(cv2.cvtColor(final_images[i], cv2.COLOR_BGR2RGB))\n    plt.axis('off')\n    if predict_classes[i].argmax() == final_labels[i].argmax() :\n        plt.title(class_mapping_dict[final_labels[i].argmax()], c='g')\n    else:\n        plt.title(class_mapping_dict[final_labels[i].argmax()], c='r')","metadata":{"execution":{"iopub.status.busy":"2023-02-01T15:29:36.587212Z","iopub.execute_input":"2023-02-01T15:29:36.587953Z","iopub.status.idle":"2023-02-01T15:29:39.246054Z","shell.execute_reply.started":"2023-02-01T15:29:36.587916Z","shell.execute_reply":"2023-02-01T15:29:39.245221Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"This work is a preliminary research for the implementation of a FCN-8s for Image Segmentation with VGG16 :\nhttps://www.kaggle.com/code/codgiw/semantic-segmentation-fcn-and-vgg16","metadata":{}}]}