{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Advanced Model Architectures\n", "> It's time to get introduced to more advanced architectures! You will create an autoencoder to reconstruct noisy images, visualize convolutional neural network activations, use deep pre-trained models to classify images and learn more about recurrent neural networks and working with text as you build a network that predicts the next word in a sentence. This is the Summary of lecture \"Introduction to Deep Learning with Keras\", via datacamp.\n", "\n", "- toc: true \n", "- badges: true\n", "- comments: true\n", "- author: Chanseok Kang\n", "- categories: [Python, Datacamp, Tensorflow-Keras, Deep_Learning]\n", "- image: images/dog.png" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import tensorflow as tf\n", "import numpy as np\n", "import pandas as pd\n", "import matplotlib.pyplot as plt\n", "\n", "plt.rcParams['figure.figsize'] = (8, 8)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Tensors, layers, and autoencoders\n", "- Tensors\n", " - main data structures used in deep learning.\n", " - Inputs/Outputs and transformations in neural networks are all presented using tensors\n", "- Autoencoders\n", "![ae](image/ae.png)\n", " - Use cases\n", " - Dimensionality reduction:\n", " - Smaller dimensional space representation of our inputs\n", " - De-noising data:\n", " - If trained with clean data, irrelevant noise will be filtered out during reconstruction\n", " - Anormaly detection\n", " - A poor reconstruction will result when the model is fed with unseen inputs" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### It's a flow of tensors\n", "If you have already built a model, you can use the `model.layers` and the `tf.keras.backend` to build functions that, provided with a valid input tensor, return the corresponding output tensor.\n", "\n", "This is a useful tool when we want to obtain the output of a network at an intermediate layer.\n", "\n", "For instance, if you get the input and output from the first layer of a network, you can build an `inp_to_out` function that returns the result of carrying out forward propagation through only the first layer for a given input tensor.\n", "\n", "So that's what you're going to do right now!" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
variaceskewnesscurtosisentropyclass
03.621608.6661-2.8073-0.446990
14.545908.1674-2.4586-1.462100
23.86600-2.63831.92420.106450
33.456609.5228-4.0112-3.594400
40.32924-4.45524.5718-0.988800
\n", "
" ], "text/plain": [ " variace skewness curtosis entropy class\n", "0 3.62160 8.6661 -2.8073 -0.44699 0\n", "1 4.54590 8.1674 -2.4586 -1.46210 0\n", "2 3.86600 -2.6383 1.9242 0.10645 0\n", "3 3.45660 9.5228 -4.0112 -3.59440 0\n", "4 0.32924 -4.4552 4.5718 -0.98880 0" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "banknote = pd.read_csv('./dataset/banknotes.csv')\n", "banknote.head()" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "from sklearn.model_selection import train_test_split\n", "\n", "X = banknote.drop(['class'], axis=1)\n", "# Normalize data\n", "X = ((X - X.mean()) / X.std()).to_numpy()\n", "y = banknote['class']\n", "\n", "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"sequential_1\"\n", "_________________________________________________________________\n", "Layer (type) Output Shape Param # \n", "=================================================================\n", "dense_2 (Dense) (None, 2) 10 \n", "_________________________________________________________________\n", "dense_3 (Dense) (None, 1) 3 \n", "=================================================================\n", "Total params: 13\n", "Trainable params: 13\n", "Non-trainable params: 0\n", "_________________________________________________________________\n" ] } ], "source": [ "from tensorflow.keras import Sequential\n", "from tensorflow.keras.layers import Dense\n", "\n", "model = Sequential()\n", "model.add(Dense(2, input_shape=(4, ), activation='relu'))\n", "model.add(Dense(1, activation='sigmoid'))\n", "model.summary()" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[array([[2.41337225e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.20011747e+00],\n", " [0.00000000e+00, 1.63594782e-01],\n", " [0.00000000e+00, 1.59020412e+00],\n", " [1.87821317e+00, 0.00000000e+00],\n", " [0.00000000e+00, 6.89649582e-01],\n", " [0.00000000e+00, 1.18970406e+00],\n", " [4.36099730e-02, 1.07809877e+00],\n", " [1.04074323e+00, 0.00000000e+00],\n", " [0.00000000e+00, 5.35320751e-02],\n", " [0.00000000e+00, 2.02383709e+00],\n", " [1.25902331e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.40000868e+00],\n", " [0.00000000e+00, 1.56177998e+00],\n", " [0.00000000e+00, 6.54872358e-02],\n", " [8.29047143e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.78345084e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 5.57610095e-01],\n", " [1.99707255e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.12326002e+00],\n", " [3.05941671e-01, 0.00000000e+00],\n", " [2.59623509e-02, 3.39980543e-01],\n", " [2.51296490e-01, 7.60058880e-01],\n", " [1.11693692e+00, 0.00000000e+00],\n", " [3.59373003e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.09660923e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.74600565e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.42224920e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [6.89101696e-01, 9.70361847e-03],\n", " [5.81549853e-02, 0.00000000e+00],\n", " [1.87910652e+00, 0.00000000e+00],\n", " [0.00000000e+00, 6.40213490e-01],\n", " [5.75123906e-01, 0.00000000e+00],\n", " [0.00000000e+00, 8.35260630e-01],\n", " [2.75657684e-01, 0.00000000e+00],\n", " [1.42686933e-01, 0.00000000e+00],\n", " [2.02093244e-01, 4.92041698e-03],\n", " [1.58853674e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.84432423e+00],\n", " [0.00000000e+00, 7.26382136e-02],\n", " [0.00000000e+00, 1.55276164e-01],\n", " [9.23556983e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [7.40955651e-01, 0.00000000e+00],\n", " [6.67462170e-01, 0.00000000e+00],\n", " [0.00000000e+00, 5.69366634e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [4.46129963e-02, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 8.84919345e-01],\n", " [0.00000000e+00, 4.88706790e-02],\n", " [4.17152226e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [3.40811685e-02, 3.10283750e-01],\n", " [4.87485439e-01, 1.69685349e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [7.44812667e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 9.42120612e-01],\n", " [1.20130338e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.76153326e+00],\n", " [0.00000000e+00, 1.87242591e+00],\n", " [2.43618563e-01, 0.00000000e+00],\n", " [0.00000000e+00, 4.31954920e-01],\n", " [5.43293953e-01, 3.34428340e-01],\n", " [0.00000000e+00, 5.05452871e-01],\n", " [1.08725083e+00, 0.00000000e+00],\n", " [4.61324573e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.22714825e-01],\n", " [3.03097039e-01, 0.00000000e+00],\n", " [0.00000000e+00, 7.16247439e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [4.33669806e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.49035692e+00],\n", " [0.00000000e+00, 1.09344757e+00],\n", " [9.16708112e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.65973628e+00],\n", " [0.00000000e+00, 2.00499352e-02],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [5.68265676e-01, 1.02209079e+00],\n", " [0.00000000e+00, 1.21419775e+00],\n", " [9.90856826e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.23398876e+00],\n", " [2.04403296e-01, 2.76350111e-01],\n", " [0.00000000e+00, 1.52938342e+00],\n", " [0.00000000e+00, 2.56740361e-01],\n", " [1.70754105e-01, 0.00000000e+00],\n", " [8.34448636e-01, 0.00000000e+00],\n", " [3.14438254e-01, 0.00000000e+00],\n", " [4.61685121e-01, 6.72646193e-03],\n", " [0.00000000e+00, 3.36706281e-01],\n", " [1.01792181e+00, 0.00000000e+00],\n", " [4.60951120e-01, 0.00000000e+00],\n", " [3.81765574e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.08976591e+00],\n", " [9.61791258e-03, 0.00000000e+00],\n", " [3.47471297e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.25504220e+00],\n", " [0.00000000e+00, 1.53693020e+00],\n", " [1.89761668e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.25547373e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.15319312e-01],\n", " [1.52636504e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.35989916e+00],\n", " [2.11954308e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.29366922e+00],\n", " [0.00000000e+00, 1.27615619e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.44746095e-01, 1.00865531e+00],\n", " [1.84737051e+00, 0.00000000e+00],\n", " [4.39038754e-01, 0.00000000e+00],\n", " [1.27399850e+00, 0.00000000e+00],\n", " [2.15075180e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.99279308e+00],\n", " [0.00000000e+00, 1.93393743e+00],\n", " [0.00000000e+00, 9.24756408e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.08459997e+00, 0.00000000e+00],\n", " [6.53292060e-01, 0.00000000e+00],\n", " [1.74425601e-03, 6.03169858e-01],\n", " [0.00000000e+00, 1.22360504e+00],\n", " [7.24477887e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.34261131e-01],\n", " [0.00000000e+00, 1.28825092e+00],\n", " [0.00000000e+00, 3.43471676e-01],\n", " [0.00000000e+00, 1.90739021e-01],\n", " [0.00000000e+00, 1.65063941e+00],\n", " [3.27075362e-01, 0.00000000e+00],\n", " [0.00000000e+00, 5.35320751e-02],\n", " [8.03765893e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.38210475e-01],\n", " [8.32469463e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.24615824e+00],\n", " [0.00000000e+00, 1.89188913e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.22369230e+00],\n", " [0.00000000e+00, 1.23862648e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 2.89220452e-01],\n", " [0.00000000e+00, 6.93475842e-01],\n", " [1.26052833e+00, 0.00000000e+00],\n", " [8.11270177e-01, 0.00000000e+00],\n", " [7.17246300e-03, 1.07612753e+00],\n", " [0.00000000e+00, 1.16010141e+00],\n", " [0.00000000e+00, 1.45201504e-01],\n", " [0.00000000e+00, 1.57141387e+00],\n", " [8.04015324e-02, 1.03055251e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.71059763e+00],\n", " [0.00000000e+00, 1.19643950e+00],\n", " [0.00000000e+00, 1.90915608e+00],\n", " [3.22056144e-01, 6.07722938e-01],\n", " [0.00000000e+00, 6.64388120e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 8.90387952e-01],\n", " [5.86072147e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.50817502e+00],\n", " [9.28590417e-01, 0.00000000e+00],\n", " [2.10868335e+00, 0.00000000e+00],\n", " [7.37695277e-01, 7.63464868e-01],\n", " [0.00000000e+00, 1.33708286e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.20783424e+00],\n", " [8.34045231e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.06116390e+00],\n", " [8.69781256e-01, 0.00000000e+00],\n", " [0.00000000e+00, 9.28749919e-01],\n", " [2.00829908e-01, 8.49336863e-01],\n", " [1.53163862e+00, 0.00000000e+00],\n", " [1.99284184e+00, 0.00000000e+00],\n", " [0.00000000e+00, 9.08511698e-01],\n", " [2.26402044e+00, 0.00000000e+00],\n", " [1.25970995e+00, 0.00000000e+00],\n", " [1.54427350e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.46838665e-01],\n", " [1.00518793e-01, 0.00000000e+00],\n", " [0.00000000e+00, 3.78883839e-01],\n", " [0.00000000e+00, 1.38945961e+00],\n", " [1.06544566e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.00382477e-01, 3.55819017e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.02865899e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.52306646e-01],\n", " [0.00000000e+00, 3.65597993e-01],\n", " [2.04810524e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [6.09301031e-01, 3.82303931e-02],\n", " [1.46752799e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.65471339e+00],\n", " [6.14384532e-01, 8.38479996e-01],\n", " [6.04060650e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.75425458e+00],\n", " [0.00000000e+00, 1.80043399e+00],\n", " [2.47199059e+00, 0.00000000e+00],\n", " [1.92456818e+00, 0.00000000e+00],\n", " [8.17747831e-01, 0.00000000e+00],\n", " [0.00000000e+00, 4.20111626e-01],\n", " [0.00000000e+00, 7.08510101e-01],\n", " [0.00000000e+00, 1.11852992e+00],\n", " [1.13794529e+00, 0.00000000e+00],\n", " [0.00000000e+00, 8.85087371e-01],\n", " [1.30358905e-01, 0.00000000e+00],\n", " [1.39758766e-01, 9.98146951e-01],\n", " [0.00000000e+00, 8.79758745e-02],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.44496882e+00],\n", " [0.00000000e+00, 8.47945333e-01],\n", " [0.00000000e+00, 1.06132634e-01],\n", " [0.00000000e+00, 1.04704106e+00],\n", " [0.00000000e+00, 7.72841275e-01],\n", " [8.40024352e-01, 7.89873421e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [4.65001643e-01, 0.00000000e+00],\n", " [7.35101759e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.88874558e-01],\n", " [0.00000000e+00, 1.89503944e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [9.98270273e-01, 0.00000000e+00],\n", " [0.00000000e+00, 5.51352203e-01],\n", " [2.62805915e+00, 0.00000000e+00],\n", " [0.00000000e+00, 8.59187067e-01],\n", " [0.00000000e+00, 5.33129990e-01],\n", " [1.56493354e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [8.55039239e-01, 1.39317721e-01],\n", " [1.26544762e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.32591283e+00],\n", " [0.00000000e+00, 1.27587032e+00],\n", " [1.81157362e+00, 0.00000000e+00],\n", " [5.32678068e-01, 0.00000000e+00],\n", " [1.94705737e+00, 0.00000000e+00],\n", " [5.06789088e-01, 0.00000000e+00],\n", " [7.77811766e-01, 0.00000000e+00],\n", " [2.53820753e+00, 0.00000000e+00],\n", " [6.41557097e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [4.60951120e-01, 0.00000000e+00],\n", " [2.22764659e+00, 0.00000000e+00],\n", " [1.91693470e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 6.98245913e-02],\n", " [7.05554605e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.14223957e+00],\n", " [0.00000000e+00, 1.86926985e+00],\n", " [1.00382477e-01, 3.55819017e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.65382552e+00],\n", " [1.70428157e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [9.82010782e-01, 3.74551892e-01],\n", " [0.00000000e+00, 9.50879790e-03],\n", " [0.00000000e+00, 6.98214248e-02],\n", " [1.85486650e+00, 0.00000000e+00],\n", " [0.00000000e+00, 5.04264534e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.41407818e-01, 9.31534529e-01],\n", " [2.52242256e-02, 1.66717798e-01],\n", " [8.55947435e-01, 0.00000000e+00],\n", " [8.66398871e-01, 0.00000000e+00],\n", " [3.00003260e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [1.43798515e-01, 8.57022285e-01],\n", " [1.04515266e+00, 0.00000000e+00],\n", " [0.00000000e+00, 2.22521830e+00],\n", " [2.38295883e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.54612207e+00],\n", " [0.00000000e+00, 5.90397000e-01],\n", " [1.79277790e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [3.86952341e-01, 0.00000000e+00],\n", " [0.00000000e+00, 4.17969227e-01],\n", " [0.00000000e+00, 7.88908839e-01],\n", " [0.00000000e+00, 1.47044230e+00],\n", " [0.00000000e+00, 6.41802132e-01],\n", " [2.73979425e-01, 0.00000000e+00],\n", " [2.03104162e+00, 0.00000000e+00],\n", " [3.89831036e-01, 0.00000000e+00],\n", " [1.97434282e+00, 0.00000000e+00],\n", " [1.96871743e-01, 4.28374588e-01],\n", " [2.67502125e-02, 1.53000608e-01],\n", " [5.77322006e-01, 0.00000000e+00],\n", " [3.32194746e-01, 4.71283555e-01],\n", " [4.07624662e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [7.28671968e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.67647406e-01],\n", " [0.00000000e+00, 7.44670391e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [6.07994199e-01, 0.00000000e+00],\n", " [3.53980780e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 8.27427506e-02],\n", " [1.06464159e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.35251784e+00],\n", " [2.06752205e+00, 0.00000000e+00],\n", " [0.00000000e+00, 5.09745836e-01],\n", " [0.00000000e+00, 1.27222002e+00],\n", " [0.00000000e+00, 2.12069556e-01],\n", " [0.00000000e+00, 8.05071235e-01],\n", " [0.00000000e+00, 1.11695266e+00],\n", " [4.79525685e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [2.50306055e-02, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.82000792e+00],\n", " [0.00000000e+00, 1.40120578e+00],\n", " [0.00000000e+00, 1.45085514e+00],\n", " [0.00000000e+00, 9.86701429e-01],\n", " [1.54390791e-02, 0.00000000e+00],\n", " [0.00000000e+00, 1.05403507e+00],\n", " [8.44364822e-01, 0.00000000e+00],\n", " [6.68035029e-03, 0.00000000e+00],\n", " [0.00000000e+00, 9.87384260e-01],\n", " [0.00000000e+00, 1.01923418e+00],\n", " [1.24934697e+00, 0.00000000e+00],\n", " [9.57717538e-01, 1.50042921e-01],\n", " [7.82056868e-01, 0.00000000e+00],\n", " [1.78352499e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 9.94785786e-01],\n", " [0.00000000e+00, 1.80472660e+00],\n", " [0.00000000e+00, 1.23860741e+00],\n", " [0.00000000e+00, 1.27492166e+00],\n", " [0.00000000e+00, 6.52845442e-01],\n", " [0.00000000e+00, 1.36377621e+00],\n", " [4.42920893e-01, 0.00000000e+00],\n", " [0.00000000e+00, 9.32486415e-01],\n", " [1.10331997e-01, 0.00000000e+00],\n", " [1.04437208e+00, 0.00000000e+00],\n", " [0.00000000e+00, 6.37910724e-01],\n", " [0.00000000e+00, 1.42864436e-01],\n", " [0.00000000e+00, 1.59217381e+00],\n", " [5.28620064e-01, 3.39544743e-01],\n", " [0.00000000e+00, 1.30564272e+00],\n", " [0.00000000e+00, 1.95538080e+00],\n", " [0.00000000e+00, 1.58302292e-01],\n", " [0.00000000e+00, 3.17262411e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.38032854e+00],\n", " [6.04809821e-01, 0.00000000e+00],\n", " [8.43918800e-01, 0.00000000e+00],\n", " [2.56839848e+00, 0.00000000e+00],\n", " [2.90803671e-01, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.22878051e+00],\n", " [0.00000000e+00, 3.72181267e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.33906448e+00],\n", " [0.00000000e+00, 2.22761005e-01],\n", " [4.14691240e-01, 0.00000000e+00],\n", " [4.30658430e-01, 5.22968769e-01],\n", " [0.00000000e+00, 1.60129592e-01],\n", " [2.34632641e-01, 0.00000000e+00],\n", " [0.00000000e+00, 1.66441774e+00],\n", " [0.00000000e+00, 6.49048090e-01],\n", " [0.00000000e+00, 7.81139791e-01],\n", " [0.00000000e+00, 1.05873775e+00],\n", " [0.00000000e+00, 1.66991925e+00],\n", " [0.00000000e+00, 1.56140316e+00],\n", " [0.00000000e+00, 1.41423297e+00],\n", " [3.10068190e-01, 0.00000000e+00],\n", " [2.21441045e-01, 0.00000000e+00],\n", " [0.00000000e+00, 4.24925119e-01],\n", " [4.75076497e-01, 0.00000000e+00],\n", " [9.17636871e-01, 0.00000000e+00],\n", " [5.41674137e-01, 0.00000000e+00],\n", " [8.99932012e-02, 0.00000000e+00],\n", " [1.64200687e+00, 0.00000000e+00],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 1.98655641e+00],\n", " [0.00000000e+00, 1.33447814e+00],\n", " [0.00000000e+00, 4.17892188e-01],\n", " [6.23544812e-01, 7.01743901e-01],\n", " [0.00000000e+00, 1.30540490e+00],\n", " [2.39228368e+00, 0.00000000e+00],\n", " [3.86015065e-02, 9.99558866e-01],\n", " [0.00000000e+00, 2.75503784e-01],\n", " [0.00000000e+00, 3.97033393e-01],\n", " [0.00000000e+00, 3.54759037e-01],\n", " [0.00000000e+00, 0.00000000e+00],\n", " [0.00000000e+00, 2.28818440e+00],\n", " [0.00000000e+00, 9.90561187e-01],\n", " [0.00000000e+00, 6.85648620e-01],\n", " [1.31632590e+00, 0.00000000e+00],\n", " [6.52459681e-01, 0.00000000e+00],\n", " [0.00000000e+00, 6.37631953e-01],\n", " [7.03163564e-01, 1.64808884e-01],\n", " [5.10819435e-01, 0.00000000e+00],\n", " [0.00000000e+00, 4.75564897e-01],\n", " [0.00000000e+00, 1.45063341e+00],\n", " [9.03004646e-01, 0.00000000e+00],\n", " [6.87576607e-02, 0.00000000e+00],\n", " [0.00000000e+00, 2.65686154e-01],\n", " [2.70112067e-01, 0.00000000e+00]], dtype=float32)]\n" ] } ], "source": [ "import tensorflow.keras.backend as K\n", "\n", "# Input tensor from the 1st layer of the model\n", "inp = model.layers[0].input\n", "\n", "# Output tensor from the 1st layer of the model\n", "out = model.layers[0].output\n", "\n", "# Define a function from inputs to outputs\n", "inp_to_out = K.function([inp], [out])\n", "\n", "# Print the results of passing X_test through the 1st layer\n", "print(inp_to_out([X_test]))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Neural separation\n", "Put on your gloves because you're going to perform brain surgery!\n", "\n", "Neurons learn by updating their weights to output values that help them better distinguish between the different output classes in your dataset. You will make use of the `inp_to_out()` function you just built to visualize the output of two neurons in the first layer of the Banknote Authentication `model` as it learns." ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "def plot():\n", " fig, axes = plt.subplots(1, 5, figsize=(16, 8))\n", " for i, a in enumerate(axes):\n", " a.scatter(layer_outputs[i][:, 0], layer_outputs[i][:, 1], c=y_test, edgecolors='none');\n", " a.set_title('Test Accuracy: {:3.1f} %'.format(float(test_accuracies[i]) * 100.));\n", " plt.tight_layout()" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [], "source": [ "model.compile(optimizer='sgd', loss='binary_crossentropy', metrics=['accuracy'])" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "412/412 [==============================] - 0s 131us/sample - loss: 0.6889 - accuracy: 0.6748\n", "412/412 [==============================] - 0s 39us/sample - loss: 0.5045 - accuracy: 0.8932\n", "412/412 [==============================] - 0s 39us/sample - loss: 0.4065 - accuracy: 0.9005\n", "412/412 [==============================] - 0s 39us/sample - loss: 0.3392 - accuracy: 0.9442\n", "412/412 [==============================] - 0s 39us/sample - loss: 0.2877 - accuracy: 0.9709\n", "412/412 [==============================] - 0s 36us/sample - loss: 0.2451 - accuracy: 0.9757\n" ] } ], "source": [ "layer_outputs = []\n", "test_accuracies = []\n", "\n", "for i in range(0, 21):\n", " # Train model for 1 epoch\n", " h = model.fit(X_train, y_train, batch_size=16, epochs=1, verbose=0)\n", " if i % 4 == 0:\n", " # Get the output of the first layer\n", " layer_outputs.append(inp_to_out([X_test])[0])\n", " \n", " # Evaluate model accuracy for this epoch\n", " test_accuracies.append(model.evaluate(X_test, y_test)[1])" ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Plot 1st vs 2nd neuron output\n", "plot()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "That took a while! If you take a look at the graphs you can see how the neurons are learning to spread out the inputs based on whether they are fake or legit dollar bills. (A single fake dollar bill is represented as a purple dot in the graph) At the start the outputs are closer to each other, the weights are learned as epochs go by so that fake and legit dollar bills get a different, further and further apart output." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Building an autoencoder\n", "Autoencoders have several interesting applications like anomaly detection or image denoising. They aim at producing an output identical to its inputs. The input will be compressed into a lower dimensional space, encoded. The model then learns to decode it back to its original form.\n", "\n", "You will encode and decode the MNIST dataset of handwritten digits, the hidden layer will encode a 32-dimensional representation of the image, which originally consists of 784 pixels (28 x 28). The autoencoder will essentially learn to turn the 784 pixels original image into a compressed 32 pixels image and learn how to use that encoded representation to bring back the original 784 pixels image." ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [], "source": [ "from tensorflow.keras.datasets import mnist\n", "\n", "(X_train, y_train), (X_test, y_test) = mnist.load_data()\n", "X_train = X_train.astype('float32') / 255.\n", "X_test = X_test.astype('float32') / 255.\n", "X_train = X_train.reshape((len(X_train), np.prod(X_train.shape[1:])))\n", "X_test = X_test.reshape((len(X_test), np.prod(X_test.shape[1:])))\n", "X_test_noise = np.load('./dataset/X_test_MNIST_noise.npy')\n", "X_test_noise = X_test_noise.reshape((len(X_test_noise), np.prod(X_test.shape[1:])))\n", "y_test_noise = np.load('./dataset/y_test_MNIST.npy')" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "> Note: When I used 'adagrad' as an optimizer, it doesn't show correct answer. But after I changed 'adam', it works." ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"autoencoder\"\n", "_________________________________________________________________\n", "Layer (type) Output Shape Param # \n", "=================================================================\n", "dense_4 (Dense) (None, 32) 25120 \n", "_________________________________________________________________\n", "dense_5 (Dense) (None, 784) 25872 \n", "=================================================================\n", "Total params: 50,992\n", "Trainable params: 50,992\n", "Non-trainable params: 0\n", "_________________________________________________________________\n" ] } ], "source": [ "# Start with a sequential model\n", "autoencoder = Sequential(name='autoencoder')\n", "\n", "# Add a dense layer with input the original image pixels and neurons the encoded representation\n", "autoencoder.add(Dense(32, input_shape=(784, ), activation='relu'))\n", "\n", "# Add an output layer with as many neurons as the original image pixels\n", "autoencoder.add(Dense(784, activation='sigmoid'))\n", "\n", "# Compile your model with adadelta\n", "autoencoder.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])\n", "\n", "# Summarize your model structure\n", "autoencoder.summary()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### De-noising like an autoencoder\n", "Okay, you have just built an `autoencoder` model. Let's see how it handles a more challenging task.\n", "\n", "First, you will build a model that encodes images, and you will check how different digits are represented with `show_encodings()`. To build the encoder you will make use of your autoencoder, that has already being trained. You will just use the first half of the network, which contains the input and the bottleneck output. That way, you will obtain a 32 number output which represents the encoded version of the input image.\n", "\n", "Then, you will apply your autoencoder to noisy images from MNIST, it should be able to clean the noisy artifacts.\n", "\n", "The digits in this noisy dataset look like this:\n", "![noise](image/noisy_mnist_sample.png)" ] }, { "cell_type": "code", "execution_count": 20, "metadata": {}, "outputs": [], "source": [ "def show_encodings(encoded_imgs,number=1):\n", " n = 5 # how many digits we will display\n", " original = X_test_noise\n", " original = original[np.where(y_test_noise == number)]\n", " encoded_imgs = encoded_imgs[np.where(y_test_noise == number)]\n", " plt.figure(figsize=(20, 4))\n", " #plt.title('Original '+str(number)+' vs Encoded representation')\n", " for i in range(min(n,len(original))):\n", " # display original imgs\n", " ax = plt.subplot(2, n, i + 1)\n", " plt.imshow(original[i].reshape(28, 28))\n", " plt.gray()\n", " ax.get_xaxis().set_visible(False)\n", " ax.get_yaxis().set_visible(False)\n", "\n", " # display encoded imgs\n", " ax = plt.subplot(2, n, i + 1 + n)\n", " plt.imshow(np.tile(encoded_imgs[i],(32,1)))\n", " plt.gray()\n", " ax.get_xaxis().set_visible(False)\n", " ax.get_yaxis().set_visible(False)\n", "\n", "def compare_plot(original,decoded_imgs):\n", " n = 4 # How many digits we will display\n", " plt.figure(figsize=(20, 4))\n", " for i in range(n):\n", " # Display original\n", " ax = plt.subplot(2, n, i + 1)\n", " plt.imshow(original[i].reshape(28, 28))\n", " plt.gray()\n", " ax.get_xaxis().set_visible(False)\n", " ax.get_yaxis().set_visible(False)\n", "\n", " # Display reconstruction\n", " ax = plt.subplot(2, n, i + 1 + n)\n", " plt.imshow(decoded_imgs[i].reshape(28, 28))\n", " plt.gray()\n", " ax.get_xaxis().set_visible(False)\n", " ax.get_yaxis().set_visible(False)\n", " plt.title('Noisy vs Decoded images')" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [], "source": [ "# Train autoencoder\n", "autoencoder.fit(X_train, X_train,\n", " epochs=100,\n", " batch_size=256,\n", " shuffle=True,\n", " validation_data=(X_test, X_test), verbose=0);" ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Build your encoder by using the first layer of your autoencoder\n", "encoder = Sequential()\n", "encoder.add(autoencoder.layers[0])\n", "\n", "# Encode the noisy images and show the encodings for your favorite number [0-9]\n", "encodings = encoder.predict(X_test_noise)\n", "show_encodings(encodings, number=1)" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "decoder_layer = autoencoder.layers[-1]\n", "decoder = tf.keras.models.Model()\n", "\n", "# Predict on the noisy images with your autoencoder\n", "decoded_imgs = autoencoder.predict(X_test)\n", "\n", "# Plot noisy vs decoded images\n", "compare_plot(X_test, decoded_imgs)" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Predict on the noisy images with your autoencoder\n", "decoded_imgs = autoencoder.predict(X_test_noise)\n", "\n", "# Plot noisy vs decoded images\n", "compare_plot(X_test_noise, decoded_imgs)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Intro to CNNs\n", "![cnn](image/cnn.png)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Building a CNN model\n", "Building a CNN model in Keras isn't much more difficult than building any of the models you've already built throughout the course! You just need to make use of convolutional layers.\n", "\n", "You're going to build a shallow convolutional `model` that classifies the MNIST digits dataset. The same one you de-noised with your autoencoder! The images are 28 x 28 pixels and just have one channel, since they are black and white pictures.\n", "\n", "Go ahead and build this small convolutional model!" ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [], "source": [ "from tensorflow.keras.layers import Conv2D, Flatten\n", "\n", "# Instantiate model\n", "model = Sequential()\n", "\n", "# Add a convolutional layer of 32 filters of size 3x3\n", "model.add(Conv2D(32, kernel_size=3, input_shape=(28, 28, 1), activation='relu'))\n", "\n", "# Add a convolutional layer of 16 filters of size 3x3\n", "model.add(Conv2D(16, kernel_size=3, activation='relu'))\n", "\n", "# Flattn the previous layer output\n", "model.add(Flatten())\n", "\n", "# Add as many outputs as classes with softmax activation\n", "model.add(Dense(10, activation='softmax'))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Looking at convolutions\n", "Inspecting the activations of a convolutional layer is a cool thing. You have to do it at least once in your lifetime!\n", "\n", "To do so, you will build a new model with the Keras Model object, which takes in a list of inputs and a list of outputs. The output you will provide to this new model is the first convolutional layer outputs when given an MNIST digit as input image.\n", "\n", "Let's look at the convolutional masks that were learned in the first convolutional layer of this model!" ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"sequential_3\"\n", "_________________________________________________________________\n", "Layer (type) Output Shape Param # \n", "=================================================================\n", "conv2d (Conv2D) (None, 26, 26, 32) 320 \n", "_________________________________________________________________\n", "conv2d_1 (Conv2D) (None, 24, 24, 16) 4624 \n", "_________________________________________________________________\n", "flatten (Flatten) (None, 9216) 0 \n", "_________________________________________________________________\n", "dense_6 (Dense) (None, 10) 92170 \n", "=================================================================\n", "Total params: 97,114\n", "Trainable params: 97,114\n", "Non-trainable params: 0\n", "_________________________________________________________________\n" ] } ], "source": [ "model.summary()" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [], "source": [ "(X_train, y_train), (X_test, y_test) = mnist.load_data()\n", "X_train = np.reshape(X_train, [-1, 28, 28, 1])\n", "X_test = np.reshape(X_test, [-1, 28, 28, 1])" ] }, { "cell_type": "code", "execution_count": 28, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Train on 60000 samples\n", "Epoch 1/30\n", "60000/60000 [==============================] - 4s 64us/sample - loss: 0.2169 - accuracy: 0.9472\n", "Epoch 2/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0722 - accuracy: 0.9785\n", "Epoch 3/30\n", "60000/60000 [==============================] - 3s 49us/sample - loss: 0.0464 - accuracy: 0.9856\n", "Epoch 4/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0348 - accuracy: 0.9893\n", "Epoch 5/30\n", "60000/60000 [==============================] - 3s 49us/sample - loss: 0.0324 - accuracy: 0.9906\n", "Epoch 6/30\n", "60000/60000 [==============================] - 3s 49us/sample - loss: 0.0233 - accuracy: 0.9930\n", "Epoch 7/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0203 - accuracy: 0.9938\n", "Epoch 8/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0206 - accuracy: 0.9946\n", "Epoch 9/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0187 - accuracy: 0.9949\n", "Epoch 10/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0151 - accuracy: 0.9958\n", "Epoch 11/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0146 - accuracy: 0.9958\n", "Epoch 12/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0158 - accuracy: 0.9962\n", "Epoch 13/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0132 - accuracy: 0.9970\n", "Epoch 14/30\n", "60000/60000 [==============================] - 3s 47us/sample - loss: 0.0138 - accuracy: 0.9966\n", "Epoch 15/30\n", "60000/60000 [==============================] - 3s 47us/sample - loss: 0.0151 - accuracy: 0.9969\n", "Epoch 16/30\n", "60000/60000 [==============================] - 3s 47us/sample - loss: 0.0155 - accuracy: 0.9967\n", "Epoch 17/30\n", "60000/60000 [==============================] - 3s 47us/sample - loss: 0.0132 - accuracy: 0.9974\n", "Epoch 18/30\n", "60000/60000 [==============================] - 3s 47us/sample - loss: 0.0110 - accuracy: 0.9978\n", "Epoch 19/30\n", "60000/60000 [==============================] - 3s 50us/sample - loss: 0.0159 - accuracy: 0.9973\n", "Epoch 20/30\n", "60000/60000 [==============================] - 3s 52us/sample - loss: 0.0153 - accuracy: 0.9974\n", "Epoch 21/30\n", "60000/60000 [==============================] - 3s 49us/sample - loss: 0.0112 - accuracy: 0.9980\n", "Epoch 22/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0139 - accuracy: 0.9976\n", "Epoch 23/30\n", "60000/60000 [==============================] - 3s 47us/sample - loss: 0.0158 - accuracy: 0.9977\n", "Epoch 24/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0142 - accuracy: 0.9979\n", "Epoch 25/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0177 - accuracy: 0.9978\n", "Epoch 26/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0106 - accuracy: 0.9986\n", "Epoch 27/30\n", "60000/60000 [==============================] - 3s 48us/sample - loss: 0.0151 - accuracy: 0.9982\n", "Epoch 28/30\n", "60000/60000 [==============================] - 3s 49us/sample - loss: 0.0144 - accuracy: 0.9984\n", "Epoch 29/30\n", "60000/60000 [==============================] - 3s 50us/sample - loss: 0.0132 - accuracy: 0.9983\n", "Epoch 30/30\n", "60000/60000 [==============================] - 3s 50us/sample - loss: 0.0185 - accuracy: 0.9980\n" ] } ], "source": [ "model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])\n", "model.fit(X_train, y_train, epochs=30, batch_size=32);" ] }, { "cell_type": "code", "execution_count": 29, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Obtain a reference to the outputs of the first layer\n", "first_layer_output = model.layers[0].output\n", "\n", "# Build a model using the model's input and the first layer output\n", "first_layer_model = tf.keras.models.Model(inputs=model.layers[0].input, outputs=first_layer_output)\n", "\n", "# Use this model to predict on X_test\n", "activations = first_layer_model.predict(X_test)\n", "\n", "fig, axs = plt.subplots(1, 3, figsize=(16, 8))\n", "\n", "# Plot the activations of first digit of X_test for the 15th filter\n", "axs[1].matshow(activations[0,:,:,14], cmap = 'viridis');\n", "\n", "# Do the same but for the 18th filter now\n", "axs[2].matshow(activations[0,:,:,17], cmap = 'viridis');\n", "\n", "axs[0].matshow(X_test[0,:,:,0], cmap='viridis');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Each neuron filter of the first layer learned a different convolution. The 15th filter (a.k.a convolutional mask) learned to detect horizontal traces in your digits. On the other hand, filter 18th seems to be checking for vertical traces." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Preparing your input image\n", "The original ResNet50 model was trained with images of size 224 x 224 pixels and a number of preprocessing operations; like the subtraction of the mean pixel value in the training set for all training images. You need to pre-process the images you want to predict on in the same way.\n", "\n", "When predicting on a single image you need it to fit the model's input shape, which in this case looks like this: (batch-size, width, height, channels),`np.expand_dims` with parameter `axis = 0` adds the batch-size dimension, representing that a single image will be passed to predict. This batch-size dimension value is 1, since we are only predicting on one image.\n", "\n", "You will go over these preprocessing steps as you prepare this dog's (named Ivy) image into one that can be classified by ResNet50.\n", "![dog](image/dog.png)" ] }, { "cell_type": "code", "execution_count": 30, "metadata": {}, "outputs": [], "source": [ "from tensorflow.keras.preprocessing import image\n", "from tensorflow.keras.applications.resnet50 import preprocess_input\n", "\n", "# Load the image with the right target size for your model\n", "img = image.load_img('./dataset/dog.png', target_size=(224, 224))\n", "\n", "# Turn it into an array\n", "img_array = image.img_to_array(img)\n", "\n", "# Expand the dimensions of the image, this is so that it fits the expected model input format\n", "img_expanded = np.expand_dims(img_array, axis=0)\n", "\n", "# Pre-process the img in the same way original images were\n", "img_ready = preprocess_input(img_expanded)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Using a real world model\n", "Okay, so Ivy's picture is ready to be used by ResNet50. It is stored in `img_ready` and now looks like this:\n", "![dogp](image/dog_processed.png)\n", "ResNet50 is a model trained on the Imagenet dataset that is able to distinguish between 1000 different labeled objects. ResNet50 is a deep model with 50 layers, you can check it in 3D [here](https://tensorspace.org/html/playground/resnet50.html).\n", "It's time to use this trained model to find out Ivy's breed!" ] }, { "cell_type": "code", "execution_count": 31, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Predicted: [('n02088364', 'beagle', 0.9073764), ('n02089867', 'Walker_hound', 0.06626676), ('n02089973', 'English_foxhound', 0.018850898)]\n" ] } ], "source": [ "from tensorflow.keras.applications.resnet50 import ResNet50, decode_predictions\n", "\n", "# Instantiate a ResNet50 model with 'imagenet' weights\n", "model = ResNet50(weights='imagenet')\n", "\n", "# Predict with ResNet50 on your already processed img\n", "preds = model.predict(img_ready)\n", "\n", "# Decode the first 3 predictions\n", "print('Predicted:', decode_predictions(preds, top=3)[0])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Custom dog image" ] }, { "cell_type": "code", "execution_count": 32, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "# Load the image with the right target size for your model\n", "img = image.load_img('./dataset/grace.jpg', target_size=(224, 224))\n", "img_array = image.img_to_array(img)\n", "img_expanded = np.expand_dims(img_array, axis=0)\n", "img_ready = preprocess_input(img_expanded)\n", "plt.imshow(img);" ] }, { "cell_type": "code", "execution_count": 33, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Predicted: [('n02085936', 'Maltese_dog', 0.17497538), ('n02098286', 'West_Highland_white_terrier', 0.12481658), ('n02086240', 'Shih-Tzu', 0.09362568)]\n" ] } ], "source": [ "# Predict with ResNet50 on your already processed img\n", "preds = model.predict(img_ready)\n", "\n", "# Decode the first 3 predictions\n", "print('Predicted:', decode_predictions(preds, top=3)[0])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Intro to LSTMs\n", "- Recurrent Neural Network (RNN)\n", "![rnn](image/rnn.png)\n", "- Long Short Term Memory (LSTM)\n", "![lstm](image/lstm.png)\n", "- When to use LSTM?\n", " - Image captioning\n", " - Speech to text\n", " - Text translation\n", " - Document summarization\n", " - Text generation\n", " - Musical composition" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Text prediction with LSTMs\n", "During the following exercises you will build a toy LSTM model that is able to predict the next word using a small text dataset. This dataset consist of cleaned quotes from the The Lord of the Ring movies.\n", "\n", "You will turn this `text` into `sequences` of length 4 and make use of the Keras `Tokenizer` to prepare the features and labels for your model!\n", "\n", "The Keras `Tokenizer` is already imported for you to use. It assigns a unique number to each unique word, and stores the mappings in a dictionary. This is important since the model deals with numbers but we later will want to decode the output numbers back into words." ] }, { "cell_type": "code", "execution_count": 34, "metadata": {}, "outputs": [], "source": [ "text = '''\n", "it is not the strength of the body but the strength of the spirit it is useless to meet revenge \n", "with revenge it will heal nothing even the smallest person can change the course of history all we have \n", "to decide is what to do with the time that is given us the burned hand teaches best after that advice about \n", "fire goes to the heart END\n", "'''" ] }, { "cell_type": "code", "execution_count": 35, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Sentences: \n", " ['it is not the', 'is not the strength', 'not the strength of', 'the strength of the', 'strength of the body'] \n", " Sequences: \n", " [[5, 3, 43, 1], [3, 43, 1, 6], [43, 1, 6, 4], [1, 6, 4, 1], [6, 4, 1, 10]]\n" ] } ], "source": [ "from tensorflow.keras.preprocessing.text import Tokenizer\n", "\n", "# Split text into an array of words\n", "words = text.split()\n", "\n", "# Make sentences of 4 words each, moving one word at a time\n", "sentences = []\n", "for i in range(4, len(words) + 1):\n", " sentences.append(' '.join(words[i - 4: i]))\n", " \n", "# Instantiate a Tokenizer, then fit it on the sentences\n", "tokenizer = Tokenizer()\n", "tokenizer.fit_on_texts(sentences)\n", "\n", "# Turn sentences into a sequence of numbers\n", "sequences = tokenizer.texts_to_sequences(sentences)\n", "print(\"Sentences: \\n {} \\n Sequences: \\n {}\".format(sentences[:5], sequences[:5]))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Build your LSTM model\n", "You've already prepared your sequences of text. It's time to build your LSTM model!\n", "\n", "Remember your sequences had 4 words each, your model will be trained on the first three words of each sequence, predicting the 4th one. You are going to use an `Embedding` layer that will essentially learn to turn words into vectors. These vectors will then be passed to a simple `LSTM` layer. Our output is a `Dense` layer with as many neurons as words in the vocabulary and `softmax` activation. This is because we want to obtain the highest probable next word out of all possible words." ] }, { "cell_type": "code", "execution_count": 36, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "46" ] }, "execution_count": 36, "metadata": {}, "output_type": "execute_result" } ], "source": [ "vocab_size = len(tokenizer.word_counts) + 1\n", "vocab_size" ] }, { "cell_type": "code", "execution_count": 37, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"sequential_4\"\n", "_________________________________________________________________\n", "Layer (type) Output Shape Param # \n", "=================================================================\n", "embedding (Embedding) (None, 3, 8) 368 \n", "_________________________________________________________________\n", "lstm (LSTM) (None, 32) 5248 \n", "_________________________________________________________________\n", "dense_7 (Dense) (None, 32) 1056 \n", "_________________________________________________________________\n", "dense_8 (Dense) (None, 46) 1518 \n", "=================================================================\n", "Total params: 8,190\n", "Trainable params: 8,190\n", "Non-trainable params: 0\n", "_________________________________________________________________\n" ] } ], "source": [ "from tensorflow.keras.layers import Embedding, LSTM\n", "\n", "model = Sequential()\n", "\n", "# Add an Embedding layer with the right parameters\n", "model.add(Embedding(input_dim=vocab_size, input_length=3, output_dim=8))\n", "\n", "# Add a 32 unit LSTM layer\n", "model.add(LSTM(32))\n", "\n", "# Add a hidden Dense layer of 32 units\n", "model.add(Dense(32, activation='relu'))\n", "model.add(Dense(vocab_size, activation='softmax'))\n", "model.summary()" ] }, { "cell_type": "code", "execution_count": 38, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(64, 4)\n" ] } ], "source": [ "from tensorflow.keras.utils import to_categorical\n", "\n", "np_sequences = np.array(sequences)\n", "print(np_sequences.shape)\n", "\n", "X = np_sequences[:, :3]\n", "y = np_sequences[:, 3]\n", "y = to_categorical(y, num_classes=vocab_size)" ] }, { "cell_type": "code", "execution_count": 39, "metadata": {}, "outputs": [], "source": [ "model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n", "history = model.fit(X, y, epochs=500, verbose=0);" ] }, { "cell_type": "code", "execution_count": 40, "metadata": {}, "outputs": [], "source": [ "def predict_text(test_text, model=model):\n", " if len(test_text.split()) != 3:\n", " print('Text input should be 3 words')\n", " return False\n", " # Turn the test_text into a sequence of numbers\n", " test_seq = tokenizer.texts_to_sequences([test_text])\n", " test_seq = np.array(test_seq)\n", " \n", " # Use the model passed as a parameter to predict the next word\n", " pred = model.predict(test_seq).argmax(axis=1)[0]\n", " \n", " # Return the word that maps to the prediction\n", " return tokenizer.index_word[pred]" ] }, { "cell_type": "code", "execution_count": 41, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "'revenge'" ] }, "execution_count": 41, "metadata": {}, "output_type": "execute_result" } ], "source": [ "predict_text('meet revenge with')" ] }, { "cell_type": "code", "execution_count": 42, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "'history'" ] }, "execution_count": 42, "metadata": {}, "output_type": "execute_result" } ], "source": [ "predict_text('the course of')" ] }, { "cell_type": "code", "execution_count": 43, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "'spirit'" ] }, "execution_count": 43, "metadata": {}, "output_type": "execute_result" } ], "source": [ "predict_text('strength of the')" ] }, { "cell_type": "code", "execution_count": 44, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "def plot_graphs(history, string):\n", " plt.plot(history.history[string])\n", " plt.xlabel('Epochs')\n", " plt.ylabel(string)\n", "\n", "plot_graphs(history, 'accuracy')" ] }, { "cell_type": "code", "execution_count": 45, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Final story : meet revenge with revenge it will heal nothing even the smallest person can change the course of history all we have to decide is what to do with the time that is given us the burned hand teaches best after that advice about fire goes to the heart\n" ] } ], "source": [ "text = 'meet revenge with'\n", "story = [text]\n", "for i in range(100):\n", " result = predict_text(text)\n", " if result == \"end\":\n", " break;\n", " story.append(result)\n", " text += ' ' + result\n", " text = ' '.join(text.split()[1:])\n", "\n", "story = \" \".join(str(x) for x in story)\n", "print('Final story : {}'.format(story))" ] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.6" } }, "nbformat": 4, "nbformat_minor": 4 }