|---|---|---|---| | [ View on TensorFlow.org](https://www.tensorflow.org/tutorials/generative/autoencoder) | [ Run in Google Colab](https://colab.research.google.com/github/tensorflow/docs/blob/master/site/en/tutorials/generative/autoencoder.ipynb) | [ View source on GitHub](https://github.com/tensorflow/docs/blob/master/site/en/tutorials/generative/autoencoder.ipynb) | [Download notebook](https://storage.googleapis.com/tensorflow_docs/docs/site/en/tutorials/generative/autoencoder.ipynb) | This tutorial introduces autoencoders with three examples: the basics, image denoising, and anomaly detection. An autoencoder is a special type of neural network that is trained to copy its input to its output. For example, given an image of a handwritten digit, an autoencoder first encodes the image into a lower dimensional latent representation, then decodes the latent representation back to an image. An autoencoder learns to compress the data while minimizing the reconstruction error. To learn more about autoencoders, please consider reading chapter 14 from [Deep Learning](https://www.deeplearningbook.org/) by Ian Goodfellow, Yoshua Bengio, and Aaron Courville. ## Import TensorFlow and other libraries import matplotlib.pyplot as plt import numpy as np import pandas as pd import tensorflow as tf from sklearn.metrics import accuracy_score, precision_score, recall_score from sklearn.model_selection import train_test_split from tensorflow.keras import layers, losses from tensorflow.keras.datasets import fashion_mnist from tensorflow.keras.models import Model ## Load the dataset To start, you will train the basic autoencoder using the Fashion MNIST dataset. Each image in this dataset is 28x28 pixels. (x_train, _), (x_test, _) = fashion_mnist.load_data() x_train = x_train.astype('float32') / 255. x_test = x_test.astype('float32') / 255. print (x_train.shape) print (x_test.shape) ``` Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz 29515/29515 [==============================] - 0s 0us/step Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz 26421880/26421880 [==============================] - 0s 0us/step Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz 5148/5148 [==============================] - 0s 0us/step Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz 4422102/4422102 [==============================] - 0s 0us/step (60000, 28, 28) (10000, 28, 28) ``` ## First example: Basic autoencoder  Define an autoencoder with two Dense layers: an `encoder`, which compresses the images into a 64 dimensional latent vector, and a `decoder`, that reconstructs the original image from the latent space. To define your model, use the [Keras Model Subclassing API](https://www.tensorflow.org/guide/keras/custom_layers_and_models). class Autoencoder(Model): def __init__(self, latent_dim, shape): super(Autoencoder, self).__init__() self.latent_dim = latent_dim self.shape = shape self.encoder = tf.keras.Sequential([ layers.Flatten(), layers.Dense(latent_dim, activation='relu'), ]) self.decoder = tf.keras.Sequential([ layers.Dense(int(tf.math.reduce_prod(shape)), activation='sigmoid'), layers.Reshape(shape) ]) def call(self, x): encoded = self.encoder(x) decoded = self.decoder(encoded) return decoded shape = x_test.shape[1:] latent_dim = 64 autoencoder = Autoencoder(latent_dim, shape) autoencoder.compile(optimizer='adam', loss=losses.MeanSquaredError()) Train the model using `x_train` as both the input and the target. The `encoder` will learn to compress the dataset from 784 dimensions to the latent space, and the `decoder` will learn to reconstruct the original images. . autoencoder.fit(x_train, x_train, epochs=10, shuffle=True, validation_data=(x_test, x_test)) ``` Epoch 1/10 1875/1875 [==============================] - 6s 3ms/step - loss: 0.0240 - val_loss: 0.0135 Epoch 2/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0118 - val_loss: 0.0107 Epoch 3/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0102 - val_loss: 0.0099 Epoch 4/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0096 - val_loss: 0.0094 Epoch 5/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0093 - val_loss: 0.0093 Epoch 6/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0091 - val_loss: 0.0092 Epoch 7/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0090 - val_loss: 0.0092 Epoch 8/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0089 - val_loss: 0.0090 Epoch 9/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0089 - val_loss: 0.0091 Epoch 10/10 1875/1875 [==============================] - 4s 2ms/step - loss: 0.0088 - val_loss: 0.0089 ``` Now that the model is trained, let's test it by encoding and decoding images from the test set. encoded_imgs = autoencoder.encoder(x_test).numpy() decoded_imgs = autoencoder.decoder(encoded_imgs).numpy() n = 10 plt.figure(figsize=(20, 4)) for i in range(n): # display original ax = plt.subplot(2, n, i + 1) plt.imshow(x_test[i]) plt.title("original") plt.gray() ax.get_xaxis().set_visible(False) ax.get_yaxis().set_visible(False) # display reconstruction ax = plt.subplot(2, n, i + 1 + n) plt.imshow(decoded_imgs[i]) plt.title("reconstructed") plt.gray() ax.get_xaxis().set_visible(False) ax.get_yaxis().set_visible(False) plt.show() ```

``` ## Second example: Image denoising  An autoencoder can also be trained to remove noise from images. In the following section, you will create a noisy version of the Fashion MNIST dataset by applying random noise to each image. You will then train an autoencoder using the noisy image as input, and the original image as the target. Let's reimport the dataset to omit the modifications made earlier. (x_train, _), (x_test, _) = fashion_mnist.load_data() x_train = x_train.astype('float32') / 255. x_test = x_test.astype('float32') / 255. x_train = x_train[..., tf.newaxis] x_test = x_test[..., tf.newaxis] print(x_train.shape) ``` (60000, 28, 28, 1) ``` Adding random noise to the images noise_factor = 0.2 x_train_noisy = x_train + noise_factor * tf.random.normal(shape=x_train.shape) x_test_noisy = x_test + noise_factor * tf.random.normal(shape=x_test.shape) x_train_noisy = tf.clip_by_value(x_train_noisy, clip_value_min=0., clip_value_max=1.) x_test_noisy = tf.clip_by_value(x_test_noisy, clip_value_min=0., clip_value_max=1.) Plot the noisy images. n = 10 plt.figure(figsize=(20, 2)) for i in range(n): ax = plt.subplot(1, n, i + 1) plt.title("original + noise") plt.imshow(tf.squeeze(x_test_noisy[i])) plt.gray() plt.show()  ### Define a convolutional autoencoder In this example, you will train a convolutional autoencoder using [Conv2D](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Conv2D) layers in the `encoder`, and [Conv2DTranspose](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Conv2DTranspose) layers in the `decoder`. class Denoise(Model): def __init__(self): super(Denoise, self).__init__() self.encoder = tf.keras.Sequential([ layers.Input(shape=(28, 28, 1)), layers.Conv2D(16, (3, 3), activation='relu', padding='same', strides=2), layers.Conv2D(8, (3, 3), activation='relu', padding='same', strides=2)]) self.decoder = tf.keras.Sequential([ layers.Conv2DTranspose(8, kernel_size=3, strides=2, activation='relu', padding='same'), layers.Conv2DTranspose(16, kernel_size=3, strides=2, activation='relu', padding='same'), layers.Conv2D(1, kernel_size=(3, 3), activation='sigmoid', padding='same')]) def call(self, x): encoded = self.encoder(x) decoded = self.decoder(encoded) return decoded autoencoder = Denoise() autoencoder.compile(optimizer='adam', loss=losses.MeanSquaredError()) autoencoder.fit(x_train_noisy, x_train, epochs=10, shuffle=True, validation_data=(x_test_noisy, x_test)) ``` Epoch 1/10 1875/1875 [==============================] - 9s 4ms/step - loss: 0.0168 - val_loss: 0.0094 Epoch 2/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0088 - val_loss: 0.0084 Epoch 3/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0081 - val_loss: 0.0081 Epoch 4/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0078 - val_loss: 0.0077 Epoch 5/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0076 - val_loss: 0.0076 Epoch 6/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0075 - val_loss: 0.0075 Epoch 7/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0074 - val_loss: 0.0074 Epoch 8/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0073 - val_loss: 0.0073 Epoch 9/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0072 - val_loss: 0.0072 Epoch 10/10 1875/1875 [==============================] - 7s 4ms/step - loss: 0.0071 - val_loss: 0.0072 ``` Let's take a look at a summary of the encoder. Notice how the images are downsampled from 28x28 to 7x7. autoencoder.encoder.summary() ``` Model: "sequential_2" _________________________________________________________________ Layer (type) Output Shape Param # ================================================================= conv2d (Conv2D) (None, 14, 14, 16) 160 conv2d_1 (Conv2D) (None, 7, 7, 8) 1160 ================================================================= Total params: 1320 (5.16 KB) Trainable params: 1320 (5.16 KB) Non-trainable params: 0 (0.00 Byte) _________________________________________________________________ ``` The decoder upsamples the images back from 7x7 to 28x28. autoencoder.decoder.summary() ``` Model: "sequential_3" _________________________________________________________________ Layer (type) Output Shape Param # ================================================================= conv2d_transpose (Conv2DTr (None, 14, 14, 8) 584 anspose) conv2d_transpose_1 (Conv2D (None, 28, 28, 16) 1168 Transpose) conv2d_2 (Conv2D) (None, 28, 28, 1) 145 ================================================================= Total params: 1897 (7.41 KB) Trainable params: 1897 (7.41 KB) Non-trainable params: 0 (0.00 Byte) _________________________________________________________________ ``` Plotting both the noisy images and the denoised images produced by the autoencoder. encoded_imgs = autoencoder.encoder(x_test_noisy).numpy() decoded_imgs = autoencoder.decoder(encoded_imgs).numpy() n = 10 plt.figure(figsize=(20, 4)) for i in range(n): # display original + noise ax = plt.subplot(2, n, i + 1) plt.title("original + noise") plt.imshow(tf.squeeze(x_test_noisy[i])) plt.gray() ax.get_xaxis().set_visible(False) ax.get_yaxis().set_visible(False) # display reconstruction bx = plt.subplot(2, n, i + n + 1) plt.title("reconstructed") plt.imshow(tf.squeeze(decoded_imgs[i])) plt.gray() bx.get_xaxis().set_visible(False) bx.get_yaxis().set_visible(False) plt.show()  ## Third example: Anomaly detection ## Overview In this example, you will train an autoencoder to detect anomalies on the [ECG5000 dataset](http://www.timeseriesclassification.com/description.php?Dataset=ECG5000). This dataset contains 5,000 [Electrocardiograms](https://en.wikipedia.org/wiki/Electrocardiography), each with 140 data points. You will use a simplified version of the dataset, where each example has been labeled either `0` (corresponding to an abnormal rhythm), or `1` (corresponding to a normal rhythm). You are interested in identifying the abnormal rhythms. > [!NOTE] > **Note:** This is a labeled dataset, so you could phrase this as a supervised learning problem. The goal of this example is to illustrate anomaly detection concepts you can apply to larger datasets, where you do not have labels available (for example, if you had many thousands of normal rhythms, and only a small number of abnormal rhythms). How will you detect anomalies using an autoencoder? Recall that an autoencoder is trained to minimize reconstruction error. You will train an autoencoder on the normal rhythms only, then use it to reconstruct all the data. Our hypothesis is that the abnormal rhythms will have higher reconstruction error. You will then classify a rhythm as an anomaly if the reconstruction error surpasses a fixed threshold. ### Load ECG data The dataset you will use is based on one from [timeseriesclassification.com](http://www.timeseriesclassification.com/description.php?Dataset=ECG5000). # Download the dataset dataframe = pd.read_csv('http://storage.googleapis.com/download.tensorflow.org/data/ecg.csv', header=None) raw_data = dataframe.values dataframe.head() # The last element contains the labels labels = raw_data[:, -1] # The other data points are the electrocadriogram data data = raw_data[:, 0:-1] train_data, test_data, train_labels, test_labels = train_test_split( data, labels, test_size=0.2, random_state=21 ) Normalize the data to `[0,1]`. min_val = tf.reduce_min(train_data) max_val = tf.reduce_max(train_data) train_data = (train_data - min_val) / (max_val - min_val) test_data = (test_data - min_val) / (max_val - min_val) train_data = tf.cast(train_data, tf.float32) test_data = tf.cast(test_data, tf.float32) You will train the autoencoder using only the normal rhythms, which are labeled in this dataset as `1`. Separate the normal rhythms from the abnormal rhythms. train_labels = train_labels.astype(bool) test_labels = test_labels.astype(bool) normal_train_data = train_data[train_labels] normal_test_data = test_data[test_labels] anomalous_train_data = train_data[~train_labels] anomalous_test_data = test_data[~test_labels] Plot a normal ECG. plt.grid() plt.plot(np.arange(140), normal_train_data[0]) plt.title("A Normal ECG") plt.show()  Plot an anomalous ECG. plt.grid() plt.plot(np.arange(140), anomalous_train_data[0]) plt.title("An Anomalous ECG") plt.show()  ### Build the model class AnomalyDetector(Model): def __init__(self): super(AnomalyDetector, self).__init__() self.encoder = tf.keras.Sequential([ layers.Dense(32, activation="relu"), layers.Dense(16, activation="relu"), layers.Dense(8, activation="relu")]) self.decoder = tf.keras.Sequential([ layers.Dense(16, activation="relu"), layers.Dense(32, activation="relu"), layers.Dense(140, activation="sigmoid")]) def call(self, x): encoded = self.encoder(x) decoded = self.decoder(encoded) return decoded autoencoder = AnomalyDetector() autoencoder.compile(optimizer='adam', loss='mae') Notice that the autoencoder is trained using only the normal ECGs, but is evaluated using the full test set. history = autoencoder.fit(normal_train_data, normal_train_data, epochs=20, batch_size=512, validation_data=(test_data, test_data), shuffle=True) ``` Epoch 1/20 5/5 [==============================] - 2s 53ms/step - loss: 0.0568 - val_loss: 0.0521 Epoch 2/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0539 - val_loss: 0.0503 Epoch 3/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0504 - val_loss: 0.0484 Epoch 4/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0468 - val_loss: 0.0461 Epoch 5/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0434 - val_loss: 0.0447 Epoch 6/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0400 - val_loss: 0.0428 Epoch 7/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0369 - val_loss: 0.0414 Epoch 8/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0342 - val_loss: 0.0401 Epoch 9/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0319 - val_loss: 0.0391 Epoch 10/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0299 - val_loss: 0.0380 Epoch 11/20 5/5 [==============================] - 0s 17ms/step - loss: 0.0282 - val_loss: 0.0374 Epoch 12/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0268 - val_loss: 0.0365 Epoch 13/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0256 - val_loss: 0.0360 Epoch 14/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0247 - val_loss: 0.0353 Epoch 15/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0240 - val_loss: 0.0347 Epoch 16/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0233 - val_loss: 0.0345 Epoch 17/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0228 - val_loss: 0.0341 Epoch 18/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0223 - val_loss: 0.0337 Epoch 19/20 5/5 [==============================] - 0s 17ms/step - loss: 0.0218 - val_loss: 0.0330 Epoch 20/20 5/5 [==============================] - 0s 18ms/step - loss: 0.0213 - val_loss: 0.0328 ``` plt.plot(history.history["loss"], label="Training Loss") plt.plot(history.history["val_loss"], label="Validation Loss") plt.legend() ``` ```  You will soon classify an ECG as anomalous if the reconstruction error is greater than one standard deviation from the normal training examples. First, let's plot a normal ECG from the training set, the reconstruction after it's encoded and decoded by the autoencoder, and the reconstruction error. encoded_data = autoencoder.encoder(normal_test_data).numpy() decoded_data = autoencoder.decoder(encoded_data).numpy() plt.plot(normal_test_data[0], 'b') plt.plot(decoded_data[0], 'r') plt.fill_between(np.arange(140), decoded_data[0], normal_test_data[0], color='lightcoral') plt.legend(labels=["Input", "Reconstruction", "Error"]) plt.show()  Create a similar plot, this time for an anomalous test example. encoded_data = autoencoder.encoder(anomalous_test_data).numpy() decoded_data = autoencoder.decoder(encoded_data).numpy() plt.plot(anomalous_test_data[0], 'b') plt.plot(decoded_data[0], 'r') plt.fill_between(np.arange(140), decoded_data[0], anomalous_test_data[0], color='lightcoral') plt.legend(labels=["Input", "Reconstruction", "Error"]) plt.show()  ### Detect anomalies Detect anomalies by calculating whether the reconstruction loss is greater than a fixed threshold. In this tutorial, you will calculate the mean average error for normal examples from the training set, then classify future examples as anomalous if the reconstruction error is higher than one standard deviation from the training set. Plot the reconstruction error on normal ECGs from the training set reconstructions = autoencoder.predict(normal_train_data) train_loss = tf.keras.losses.mae(reconstructions, normal_train_data) plt.hist(train_loss[None,:], bins=50) plt.xlabel("Train loss") plt.ylabel("No of examples") plt.show() ``` 74/74 [==============================] - 0s 1ms/step ```  Choose a threshold value that is one standard deviations above the mean. threshold = np.mean(train_loss) + np.std(train_loss) print("Threshold: ", threshold) ``` Threshold: 0.03243865 ``` > [!NOTE] > **Note:** There are other strategies you could use to select a threshold value above which test examples should be classified as anomalous, the correct approach will depend on your dataset. You can learn more with the links at the end of this tutorial. If you examine the reconstruction error for the anomalous examples in the test set, you'll notice most have greater reconstruction error than the threshold. By varing the threshold, you can adjust the [precision](https://developers.google.com/machine-learning/glossary#precision) and [recall](https://developers.google.com/machine-learning/glossary#recall) of your classifier. reconstructions = autoencoder.predict(anomalous_test_data) test_loss = tf.keras.losses.mae(reconstructions, anomalous_test_data) plt.hist(test_loss[None, :], bins=50) plt.xlabel("Test loss") plt.ylabel("No of examples") plt.show() ``` 14/14 [==============================] - 0s 2ms/step ```  Classify an ECG as an anomaly if the reconstruction error is greater than the threshold. def predict(model, data, threshold): reconstructions = model(data) loss = tf.keras.losses.mae(reconstructions, data) return tf.math.less(loss, threshold) def print_stats(predictions, labels): print("Accuracy = {}".format(accuracy_score(labels, predictions))) print("Precision = {}".format(precision_score(labels, predictions))) print("Recall = {}".format(recall_score(labels, predictions))) preds = predict(autoencoder, test_data, threshold) print_stats(preds, test_labels) ``` Accuracy = 0.943 Precision = 0.9921722113502935 Recall = 0.9053571428571429 ``` ## Next steps To learn more about anomaly detection with autoencoders, check out this excellent [interactive example](https://anomagram.fastforwardlabs.com/#/) built with TensorFlow.js by Victor Dibia. For a real-world use case, you can learn how [Airbus Detects Anomalies in ISS Telemetry Data](https://blog.tensorflow.org/2020/04/how-airbus-detects-anomalies-iss-telemetry-data-tfx.html) using TensorFlow. To learn more about the basics, consider reading this [blog post](https://blog.keras.io/building-autoencoders-in-keras.html) by François Chollet. For more details, check out chapter 14 from [Deep Learning](https://www.deeplearningbook.org/) by Ian Goodfellow, Yoshua Bengio, and Aaron Courville.