{"cells":[{"metadata":{"trusted":true},"cell_type":"code","source":"import pandas as pd\nimport os\nimport numpy as np\nimport PIL.Image\nimport glob\n\nimport tensorflow\nfrom tensorflow.keras import Sequential\nfrom tensorflow.keras.layers import *\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator\n\nfrom sklearn.model_selection import train_test_split\n\nfrom matplotlib import pyplot as plt\nfrom matplotlib import rcParams\nimport seaborn as sns\nimport cv2\nimport tqdm\nfrom IPython.display import Image\n%matplotlib inline\n\nos.listdir(\"../input/cifar-10\")","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Importing training labels"},{"metadata":{"trusted":true},"cell_type":"code","source":"!pip install py7zr","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"!python -m py7zr x ../input/cifar-10/train.7z ","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"y = pd.read_csv('../input/cifar-10/trainLabels.csv')\nfile_list = glob.glob(\"../working/train/*\")","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"y","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Extracting Image Names"},{"metadata":{"trusted":true},"cell_type":"code","source":"file_list = sorted(file_list,key= lambda x:int(x.split(\"/\")[-1].split(\".\")[0]))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Forming our training X"},{"metadata":{"trusted":true},"cell_type":"code","source":"X = np.array([np.array(PIL.Image.open(f_name)) for f_name in file_list])\nprint(f\"{X.shape} -- {y.shape}\")","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"pd.CategoricalIndex(y['label'].values).categories","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"temp = y['label'].values\nX = X.astype('float32')\n#Normalizing\nX = X/255.0\ncategories = list(pd.CategoricalIndex(temp).categories)\nmapped = {}\nrev_mapped = {}\nfor i in range(len(categories)):\n    mapped[categories[i]] = i\n    rev_mapped[i] = categories[i]\n\nfor i in range(len(temp)):\n    temp[i] = mapped[temp[i]]\n    \n#y is the one-hot-vector for each datapoint/image\ny = tensorflow.keras.utils.to_categorical(temp)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(f\"{X.shape} -- {y.shape}\")","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Model Construction\n"},{"metadata":{"trusted":true},"cell_type":"code","source":"\nmodel = Sequential()\nmodel.add(Conv2D(24,kernel_size=(3,3),padding='same',activation='relu',input_shape=(32,32,3)))\nmodel.add(BatchNormalization())\nmodel.add(Conv2D(24,kernel_size=(3,3),padding='same',activation='relu'))\nmodel.add(BatchNormalization())\nmodel.add(Conv2D(32,kernel_size=5,strides=2,padding='same',activation='relu'))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.4))\nmodel.add(Conv2D(64,(3,3),activation='relu'))\nmodel.add(BatchNormalization())\nmodel.add(Conv2D(64,kernel_size=5,strides=2,padding='same',activation='relu'))\nmodel.add(BatchNormalization())\nmodel.add(Dropout(0.4))\nmodel.add(Conv2D(128,kernel_size=4,activation='relu'))\nmodel.add(BatchNormalization())\nmodel.add(Flatten())\nmodel.add(Dropout(0.4))\nmodel.add(Dense(10,activation='softmax'))\n\nmodel.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy'])\n\n\n\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model.summary()","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Data Augmentation\n"},{"metadata":{"trusted":true},"cell_type":"code","source":"datagen = ImageDataGenerator(rotation_range=10, zoom_range=0.1, width_shift_range=0.1, height_shift_range=0.1)\ndatagen.fit(X)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Splitting dataset into training and testing"},{"metadata":{"trusted":true},"cell_type":"code","source":"X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=123)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Training our CNN Model"},{"metadata":{"trusted":true},"cell_type":"code","source":"history = model.fit(datagen.flow(X_train,y_train),batch_size=128,epochs=40,verbose=2,validation_data=(X_test,y_test))","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## We have 85% accuracy with this simple CNN model!!"},{"metadata":{"trusted":true},"cell_type":"code","source":"acc = history.history['accuracy']\nval_acc = history.history['val_accuracy']\nloss = history.history['loss']\nval_loss = history.history['val_loss']\n\nepochs=range(len(acc))\n\nplt.plot(epochs,acc, )\nplt.plot(epochs,val_acc)\nplt.title('Training and validation accuracy')\nplt.legend([\"training accuracy\", \"validation accuracy\"], loc =\"lower right\") \n\nplt.figure()\n\nplt.xlabel(\"Epochs\")\nplt.plot(epochs,loss)\nplt.plot(epochs,val_loss)\nplt.title('Training and validation loss')\nplt.legend([\"training loss\", \"validation loss\"], loc =\"upper right\") ","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"# Some Missclassifications"},{"metadata":{},"cell_type":"markdown","source":""},{"metadata":{"trusted":true},"cell_type":"code","source":"pred = model.predict(np.expand_dims(X_train[10], axis=0))\nplt.figure(figsize = (1,1))\nplt.imshow(X_train[10])\nplt.xlabel(rev_mapped[np.argmax(pred, axis=1)[0]])\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"pred = model.predict(np.expand_dims(X_train[1], axis=0))\nplt.figure(figsize = (1,1))\nplt.imshow(X_train[1])\nplt.xlabel(rev_mapped[np.argmax(pred, axis=1)[0]])\n","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Some Missclassifications"},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"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":4,"nbformat_minor":4}