Skip to content

Instantly share code, notes, and snippets.

@ahlusar1989
Last active June 9, 2020 18:31
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save ahlusar1989/8a454fb254bdc23d66a49c11c1009b46 to your computer and use it in GitHub Desktop.
Save ahlusar1989/8a454fb254bdc23d66a49c11c1009b46 to your computer and use it in GitHub Desktop.
Display the source blob
Display the rendered blob
Raw
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Expectation maximization to stochastic variational inference\n",
"\n",
"This is Saran Ahluwalia's rendition of variational auto-encoders.\n",
"\n",
"## Introduction\n",
"\n",
"Given a probabilistic model $p(\\mathbf{x};\\boldsymbol\\theta)$ and some observations $\\mathbf{x}$, we often want to estimate optimal parameter values $\\boldsymbol{\\hat{\\theta}}$ that maximize the data likelihood. This can be done via [maximum likelihood](https://en.wikipedia.org/wiki/Maximum_likelihood_estimation) (ML) estimation or [maximum a posteriori](https://de.wikipedia.org/wiki/Maximum_a_posteriori) (MAP) estimation if point estimates of $\\boldsymbol\\theta$ are sufficient:\n",
"\n",
"$$\n",
"\\boldsymbol{\\hat{\\theta}} = \\underset{\\boldsymbol\\theta}{\\mathrm{argmax}}\\ p(\\mathbf{x};\\boldsymbol\\theta)\\tag{1}\n",
"$$\n",
"\n",
"In many cases, direct computation and optimization of the likelihood function $p(\\mathbf{x};\\boldsymbol\\theta)$ is complex or impossible. One option to ease computation is the introduction of [latent variables](https://en.wikipedia.org/wiki/Latent_variable) $\\mathbf{t}$ so that we have a complete data likelihood $p(\\mathbf{x},\\mathbf{t};\\boldsymbol\\theta)$ which can be decomposed into a conditional likelihood $p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)$ and a prior $p(\\mathbf{t})$.\n",
"\n",
"$$\n",
"p(\\mathbf{x},\\mathbf{t};\\boldsymbol\\theta) = p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)p(\\mathbf{t})\\tag{2}\n",
"$$\n",
"\n",
"Latent variables are not observed directly but assumed to cause observations $\\mathbf{x}$. Their choice is problem-dependent. To obtain the the marginal likelihood $p(\\mathbf{x};\\boldsymbol\\theta)$, we have to integrate i.e. marginalize out the latent variables.\n",
"\n",
"$$\n",
"p(\\mathbf{x};\\boldsymbol\\theta) = \n",
"\\int p(\\mathbf{x},\\mathbf{t};\\boldsymbol\\theta)d\\mathbf{t} = \n",
"\\int p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)p(\\mathbf{t})d\\mathbf{t}\n",
"\\tag{3}\n",
"$$\n",
"\n",
"Usually, we choose a latent variable model such that parameter estimation for the conditional likelihood $p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)$ is easier than for the marginal likelihood $p(\\mathbf{x};\\boldsymbol\\theta)$. For example, the conditional likelihood of a Gaussian [mixture model](https://en.wikipedia.org/wiki/Mixture_model) (GMM) is a single Gaussian for which parameter estimation is easier than for the marginal likelihood which is a mixture of Gaussians. The latent variable $\\mathbf{t}$ in a GMM determines the soft-assignment to mixture components and follows a categorical distribution. If we can solve the integral in Eq. 3 we can also compute the posterior distribution of the latent variables by using [Bayes' theorem](https://en.wikipedia.org/wiki/Bayes%27_theorem):\n",
"\n",
"$$\n",
"p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta) = \n",
"\\frac{p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)p(\\mathbf{t})}\n",
" {p(\\mathbf{x};\\boldsymbol\\theta)}\n",
"\\tag{4}\n",
"$$\n",
"\n",
"With the posterior, inference for the latent variables becomes possible. Note that in this article the term *estimation* is used to refer to (point) estimation of parameters via ML or MAP and *inference* to refer to Bayesian inference of random variables by computing the posterior. \n",
"\n",
"A major challenge in Bayesian inference is that the integral in Eq. 3 is often impossible or very difficult to compute in closed form. Therefore, many techniques exist to approximate the posterior in Eq. 4. They can be classified into numerical approximations ([Monte Carlo techniques](https://en.wikipedia.org/wiki/Monte_Carlo_method)) and deterministic approximations. This article is about deterministic approximations only, and their stochastic variants.\n",
"\n",
"## Expectation maximization (EM)\n",
"\n",
"Basis for many inference methods is the [expectation-maximization](https://en.wikipedia.org/wiki/Expectation-maximization_algorithm) (EM) algorithm. It is an iterative algorithm for estimating the parameters of latent variable models, often with closed-form updates at each step. We start with a rather general view of the EM algorithm that also serves as a basis for discussing variational inference methods later. It is straightforward to show<sup>[2]</sup> that the marginal log likelihood can be written as \n",
"\n",
"$$\n",
"\\log p(\\mathbf{x};\\boldsymbol\\theta) = \n",
" \\mathcal{L}(q, \\boldsymbol\\theta) + \n",
" \\mathrm{KL}(q \\mid\\mid p)\n",
"\\tag{5}\n",
"$$\n",
"\n",
"with \n",
"\n",
"$$\n",
"\\mathcal{L}(q, \\boldsymbol\\theta) = \\int q(\\mathbf{t}) \\log \n",
" \\frac{p(\\mathbf{x},\\mathbf{t};\\boldsymbol\\theta)}\n",
" {q(\\mathbf{t})} d\\mathbf{t}\n",
"\\tag{6}\n",
"$$\n",
"\n",
"and \n",
"\n",
"$$\n",
"\\mathrm{KL}(q \\mid\\mid p) = - \\int q(\\mathbf{t}) \\log \n",
" \\frac{p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)}\n",
" {q(\\mathbf{t})} d\\mathbf{t}\n",
"\\tag{7}\n",
"$$\n",
"\n",
"where $q(\\mathbf{t})$ is any probability density function. $\\mathrm{KL}(q \\mid\\mid p)$ is the [Kullback-Leibler divergence](https://en.wikipedia.org/wiki/Kullback%E2%80%93Leibler_divergence) between $q(\\mathbf{t})$ and $p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)$ that measures how much $q$ diverges from $p$. The Kullback-Leibler divergence is zero for identical distributions and greater than zero otherwise. Thus, $\\mathcal{L}(q, \\boldsymbol\\theta)$ is a lower bound of the log likelihood. It is equal to the log likelihood if $q(\\mathbf{t}) = p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)$. In the E-step of the EM algorithm, $q(\\mathbf{t})$ is therefore set to $p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)$ using the parameter values of the previous iteration $l-1$. \n",
"\n",
"$$\n",
"q^{l}(\\mathbf{t}) = p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta^{l-1})\\tag{8}\n",
"$$\n",
"\n",
"Note that this requires that $p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)$ is known, like in the GMM case where the posterior is a categorical distribution, as mentioned above. In the M-step, $\\mathcal{L}(q, \\boldsymbol\\theta)$ is optimized w.r.t. $\\boldsymbol\\theta$ using $q(\\mathbf{t})$ from the E-step:\n",
"\n",
"$$\n",
"\\boldsymbol\\theta^{l} = \\underset{\\boldsymbol\\theta}{\\mathrm{argmax}}\\ \\mathcal{L}(q^{l}, \\boldsymbol\\theta)\\tag{9}\n",
"$$\n",
"\n",
"In general, this is much simpler than optimizing $p(\\mathbf{x};\\boldsymbol\\theta)$ directly. E and M steps are repeated until convergence. However, the requirement that the posterior $p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)$ must be known is rather restrictive and there are many cases where the posterior is intractable. In these cases, further approximations must be made.\n",
"\n",
"## Variational EM\n",
"\n",
"If the posterior is unknown, we have to assume specific forms of $q(\\mathbf{t})$ and maximize the lower bound $\\mathcal{L}(q, \\boldsymbol\\theta)$ w.r.t. these functions. The area of mathematics related to these optimization problems is called [calculus of variations](https://en.wikipedia.org/wiki/Calculus_of_variations)<sup>[3]</sup>, hence the name *variational EM*, or *variationial inference* in general. A widely used approximation for the unknown posterior is the [mean-field approximation](https://en.wikipedia.org/wiki/Mean_field_theory)<sup>[2][3]</sup> which factorizes $q(\\mathbf{t})$ into $M$ partitions:\n",
"\n",
"$$\n",
"q(\\mathbf{t}) = \\prod_{i=1}^{M} q_i(\\mathbf{t}_i)\\tag{10}\n",
"$$\n",
"\n",
"For example, if $\\mathbf{t}$ is 10-dimensional, we can factorize $q(\\mathbf{t})$ into a product of 10 $q_i(\\mathbf{t}_i)$, one for each dimension, assuming independence between dimensions. The approximate posterior $q(\\mathbf{t})$ can be obtained by minimizing $\\mathrm{KL}(q \\mid\\mid \\tilde{p})$ which is the KL divergence between the factorized distribution $q(\\mathbf{t})$ and the *unnormalized* posterior $\\tilde{p}(\\mathbf{t};\\boldsymbol\\theta) = p(\\mathbf{x},\\mathbf{t};\\boldsymbol\\theta)$. This leads to the following update formula for $q_i(\\mathbf{t}_i)$:\n",
"\n",
"$$\n",
"q_i^l(\\mathbf{t}_i) = \\mathbb{E}_{-q_i} \\left[ \\log\\tilde{p}(\\mathbf{t};\\boldsymbol\\theta^{l-1}) \\right] + \\mathrm{const} \\tag{11}\n",
"$$\n",
"\n",
"where $\\mathbb{E}_{-q_i}$ denotes the expectation with respect to all variables of $\\mathbf{t}$ except $t_i$. $\\boldsymbol\\theta^{l-1}$ are the parameters from the previous iteration. This is repeated for all $q_i$ until convergence. The E-step of the variational EM algorithm is therefore\n",
"\n",
"$$\n",
"q^l(\\mathbf{t}) = \\prod_{i=1}^{M} q_i^l(\\mathbf{t}_i)\\tag{12}\n",
"$$\n",
"\n",
"and the M-step uses the posterior approximation $q^l(\\mathbf{t})$ from the E-step to estimate parameters $\\boldsymbol\\theta^l$:\n",
"\n",
"$$\n",
"\\boldsymbol\\theta^{l} = \\underset{\\boldsymbol\\theta}{\\mathrm{argmax}}\\ \\mathcal{L}(q^{l}, \\boldsymbol\\theta)\\tag{13}\n",
"$$\n",
"\n",
"The mean field approach allows inference for many interesting latent variable models but it requires analytical solutions for the approximate posterior which is not always possible. Especially when used in context of deep learning where the approximate posterior $q(\\mathbf{t})$ and the conditional likelihood $p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)$ are neural networks with at least one non-linear hidden layer, the mean field approach is not applicable any more<sup>[4]</sup>. Further approximations are required.\n",
"\n",
"## Stochastic variational inference\n",
"\n",
"Let's assume we have a latent variable model with one latent variable $\\mathbf{t}^{(i)}$ for each observation $\\mathbf{x}^{(i)}$. Observations $\\mathbf{x}^{(i)}$ come from an i.i.d. dataset. To make the following more concrete let's say that $\\mathbf{x}^{(i)}$ are images and $\\mathbf{t}^{(i)}$ are $D$-dimensional latent vectors that cause the generation of $\\mathbf{x}^{(i)}$ under the generative model $p(\\mathbf{x},\\mathbf{t};\\boldsymbol\\theta) = p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)p(\\mathbf{t})$.\n",
"\n",
"Our goal is to find optimal parameter values for the marginal likelihood $p(\\mathbf{x};\\boldsymbol\\theta)$ by maximizing its variational lower bound. Here, we neither know the true posterior $p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)$ nor can we apply the mean field approximation<sup>[4]</sup>, so we have to make further approximations. We start by assuming that $q(\\mathbf{t})$ is a factorized Gaussian i.e. a Gaussian with a diagonal covariance matrix and that we have a separate distribution $q^{(i)}$ for each latent variable $\\mathbf{t}^{(i)}$: \n",
"\n",
"$$\n",
"q^{(i)}(\\mathbf{t}^{(i)}) = \n",
" \\mathcal{N}(\\mathbf{t}^{(i)} \\lvert \\mathbf{m}^{(i)},\\mathrm{diag}(\\mathbf{s}^{2(i)}))\n",
"\\tag{14}\n",
"$$\n",
"\n",
"The problem here is that we have to estimate too many parameters. For example, if the latent space is 50-dimensional we have to estimate about 100 parameters per training object! This is not what we want. Another option is that all $q^{(i)}$ share their parameters $\\mathbf{m}$ and $\\mathbf{s}^2$ i.e. all $q^{(i)}$ are identical. This would keep the number of parameters constant but would be too restrictive though. If we want to support different $q^{(i)}$ for different $\\mathbf{t}^{(i)}$ but with a limited number of parameters we should consider using parameters for $q$ that are functions of $\\mathbf{x}^{(i)}$. These functions are themselves parametric functions that share a set of parameters $\\boldsymbol\\phi$:\n",
"\n",
"$$\n",
"q^{(i)}(\\mathbf{t}^{(i)}) = \\mathcal{N}(\\mathbf{t}^{(i)} \\lvert \n",
" m(\\mathbf{x}^{(i)},\\boldsymbol\\phi), \\mathrm{diag}(s^2(\\mathbf{x}^{(i)},\\boldsymbol\\phi)))\n",
"\\tag{15}\n",
"$$\n",
"\n",
"So we finally have a variational distribution $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ with a fixed number of parameters $\\boldsymbol\\phi$ as approximation for the true but unknown posterior $p(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\theta)$. To implement the (complex) functions $m$ and $s$ that map from an input image to the mean and the variance of that distribution we can use a [convolutional neural network](https://de.wikipedia.org/wiki/Convolutional_Neural_Network) (CNN) that is parameterized by $\\boldsymbol\\phi$. Similarly, for implementing $p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)$ we can use another neural network, parameterized by $\\boldsymbol\\theta$, that maps a latent vector $\\mathbf{t}$ to the sufficient statistics of that probability distribution. Since $\\mathbf{t}$ is often a lower-dimensional embedding or code of image $\\mathbf{x}$, $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ is referred to as *probabilistic encoder* and $p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)$ as *probabilistic decoder*.\n",
"\n",
"![vae-1](images/vae/auto-encoder-1.png)\n",
"\n",
"### Variational auto-encoder\n",
"\n",
"Both, encoder and decoder, can be combined to a *variational auto-encoder*<sup>[4]</sup> that is trained with the variational lower bound $\\mathcal{L}$ as optimization objective using standard stochastic gradient ascent methods. For our model, the variational lower bound for a single training object $\\mathbf{x}^{(i)}$ can also be formulated as:\n",
"\n",
"$$\n",
"\\mathcal{L}(\\boldsymbol\\theta, \\boldsymbol\\phi, \\mathbf{x}^{(i)}) =\n",
" \\mathbb{E}_{q(\\mathbf{t} \\lvert \\mathbf{x}^{(i)};\\boldsymbol\\phi)} \\left[\\log p(\\mathbf{x}^{(i)} \\lvert \\mathbf{t};\\boldsymbol\\theta)\\right] - \n",
" \\mathrm{KL}(q(\\mathbf{t} \\lvert \\mathbf{x}^{(i)};\\boldsymbol\\phi) \\mid\\mid p(\\mathbf{t}))\n",
"\\tag{16}\n",
"$$\n",
"\n",
"The first term is the expected negative *reconstruction error* of an image $\\mathbf{x}^{(i)}$. This term is maximized when the reconstructed image is as close as possible to the original image. It is computed by first feeding an input image $\\mathbf{x}^{(i)}$ through the encoder to compute the mean and the variance of the variational distribution $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$. To compute an approximate value of the expected negative reconstruction error, we sample from the variational distribution. Since this is a Gaussian distribution, sampling is very efficient. To compute $p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)$ we feed the samples through the decoder. A single sample per training object is usually sufficient<sup>[4]</sup> if the mini-batch size during training is large enough e.g. > 100.\n",
"\n",
"![vae-2](images/vae/auto-encoder-2.png)\n",
"\n",
"The second term in Eq. 16, the negative KL divergence, is maximized when the approximate posterior $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ is equal to the prior $p(\\mathbf{t})$. The prior is usually chosen to be the standard normal distribution $\\mathcal{N}(\\mathbf{0},\\mathbf{I})$. This term therefore acts as a regularization term to avoid that the variance of $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ becomes zero, otherwise, $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ would degenerate to a delta function and the variational auto-encoder to a usual auto-encoder. Regularizing $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ to have non-zero variance makes the decoder more robust against small changes in $\\mathbf{t}$ and the latent space a continuous space of codes that can be decoded to realistic images. \n",
"\n",
"### Gradient of the variational lower bound\n",
"\n",
"To be able to use the variational lower bound as optimization objective or loss function in tools like [Tensorflow](https://www.tensorflow.org/), we have to make sure that it is differentiable. This is easy to achieve for the regularization term which can be integrated analytically in the Gaussian case\n",
"\n",
"$$- \\mathrm{KL}(q(\\mathbf{t} \\lvert \\mathbf{x}^{(i)};\\boldsymbol\\phi) \\mid\\mid p(\\mathbf{t})) =\n",
" \\frac{1}{2} \\sum_{j=1}^{D}(1 + \\log((s_j)^2) - (m_j)^2 - (s_j)^2)\n",
"\\tag{17}\n",
"$$\n",
"\n",
"where $m_j$ and $s_j$ denote the $j$-th elements of the vectors computed with functions $m$ and $s$ (see Eq. 15). $D$ is the dimensionality of these vectors. The computation of the expected negative reconstruction error, on the other hand, involves sampling from $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ which is not differentiable. However, a simple reparameterization trick allows to express the random variable $\\mathbf{t}$ as deterministic variable $\\mathbf{t} = g(\\mathbf{m}, \\mathbf{s}, \\mathbf\\epsilon) = \\mathbf{m} + \\mathbf{s} \\mathbf\\epsilon$ plus a random (noise) variable $\\mathbf\\epsilon \\sim \\mathcal{N}(\\mathbf{0},\\mathbf{I})$ that doesn't depend on any parameters to be optimized. With this trick we can easily compute the gradient of function $g$ and can ignore $\\mathbf\\epsilon$ i.e. the sampling procedure during back-propagation.\n",
"\n",
"![vae-3](images/vae/auto-encoder-3.png)\n",
"\n",
"We haven't defined the functional form of the probabilistic decoder $p(\\mathbf{x} \\lvert \\mathbf{t};\\boldsymbol\\theta)$ yet. If we train the variational auto-encoder with grey-scale [MNIST images](https://en.wikipedia.org/wiki/MNIST_database), for example, it makes sense to use a multivariate Bernoulli distribution. In this case, the output of the decoder network is the single parameter of this distribution. It defines for each pixel the probability of being white. These probabilities are then simply mapped to values from 0-255 to generate grey-scale images. In the output layer of the decoder network there is one node per pixel with a sigmoid activation function. Hence, we compute the binary cross-entropy between the input image and the decoder output to estimate the expected reconstruction error.\n",
"\n",
"Stochastic variational inference algorithms implemented as variational auto-encoders scale to very large datasets as they can be trained based on mini-batches. Furthermore, they can also be used for data other than image data. For example, Gómez-Bombarelli et. al.<sup>[5]</sup> use a sequential representation of chemical compounds together with an RNN-based auto-encoder to infer a continuous latent space of chemical compounds that can be used e.g. for generating new chemical compounds with properties that are desirable for drug discovery. I'll cover this together with an implementation example of a variational auto-encoder."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Implementation\n",
"\n",
"This section provides an example implementation of a variational auto-encoder in [Keras](https://keras.io/) that is trained with the [MNIST handwritten digits dataset](https://en.wikipedia.org/wiki/MNIST_database)."
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Using TensorFlow backend.\n"
]
}
],
"source": [
"import numpy as np\n",
"\n",
"import matplotlib.pyplot as plt\n",
"from matplotlib import cm\n",
"\n",
"import keras\n",
"from keras import backend as K\n",
"from keras import layers\n",
"from keras.datasets import mnist\n",
"from keras.models import Model, Sequential\n",
"from keras.utils import to_categorical\n",
"\n",
"%matplotlib inline"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The architecture of the encoder and decoder network was taken from \\[6\\] and from the [accompanying notebook](https://github.com/fchollet/deep-learning-with-python-notebooks/blob/master/8.4-generating-images-with-vaes.ipynb). Here, we choose a 2-dimensional latent space for easier visualization. Reconstruction quality of images can be increased by choosing a higher-dimensional latent space and/or by using encoder and decoder models with higher capacity."
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"# Dimensions of MNIST images \n",
"image_shape = (28, 28, 1)\n",
"\n",
"# Dimension of latent space\n",
"latent_dim = 2\n",
"\n",
"# Mini-batch size for training\n",
"batch_size = 128\n",
"\n",
"def create_encoder():\n",
" '''\n",
" Creates a convolutional encoder model for MNIST images.\n",
" \n",
" - Input for the created model are MNIST images. \n",
" - Output of the created model are the sufficient statistics\n",
" of the variational distriution q(t|x;phi), mean and log \n",
" variance. \n",
" '''\n",
" encoder_iput = layers.Input(shape=image_shape)\n",
" \n",
" x = layers.Conv2D(32, 3, padding='same', activation='relu')(encoder_iput)\n",
" x = layers.Conv2D(64, 3, padding='same', activation='relu', strides=(2, 2))(x)\n",
" x = layers.Conv2D(64, 3, padding='same', activation='relu')(x)\n",
" x = layers.Conv2D(64, 3, padding='same', activation='relu')(x)\n",
" x = layers.Flatten()(x)\n",
" x = layers.Dense(32, activation='relu')(x)\n",
"\n",
" t_mean = layers.Dense(latent_dim)(x)\n",
" t_log_var = layers.Dense(latent_dim)(x)\n",
"\n",
" return Model(encoder_iput, [t_mean, t_log_var], name='encoder')\n",
"\n",
"def create_decoder():\n",
" '''\n",
" Creates a (de-)convolutional decoder model for MNIST images.\n",
" \n",
" - Input for the created model are latent vectors t.\n",
" - Output of the model are images of shape (28, 28, 1) where\n",
" the value of each pixel is the probability of being white.\n",
" '''\n",
" decoder_input = layers.Input(shape=(latent_dim,))\n",
" \n",
" x = layers.Dense(12544, activation='relu')(decoder_input)\n",
" x = layers.Reshape((14, 14, 64))(x)\n",
" x = layers.Conv2DTranspose(32, 3, padding='same', activation='relu', strides=(2, 2))(x)\n",
" x = layers.Conv2D(1, 3, padding='same', activation='sigmoid')(x)\n",
" \n",
" return Model(decoder_input, x, name='decoder')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The encoder model is trained to generate the sufficient statistics of the variational distribution $q(\\mathbf{t} \\lvert \\mathbf{x};\\boldsymbol\\phi)$ but instead of generating the standard deviation $\\mathbf{s}$ directly, as described above, it is trained to generate $\\log \\mathbf{s}^2$ as this is easier to learn, for numerical reasons.\n",
"\n",
"To implement the noise variable $\\mathbf\\epsilon$ and function g (see *reparameterization* in [Gradient of the variational lower bound](#Gradient-of-the-variational-lower-bound)), we implement a separate sampling layer that depends on the sufficient statistics of the variational distribution.\n"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"def sample(args):\n",
" '''\n",
" Draws samples from a standard normal and scales the samples with\n",
" standard deviation of the variational distribution and shifts them\n",
" by the mean.\n",
" \n",
" Args:\n",
" args: sufficient statistics of the variational distribution.\n",
" \n",
" Returns:\n",
" Samples from the variational distribution.\n",
" '''\n",
" t_mean, t_log_var = args\n",
" t_sigma = K.sqrt(K.exp(t_log_var))\n",
" epsilon = K.random_normal(shape=K.shape(t_mean), mean=0., stddev=1.)\n",
" return t_mean + t_sigma * epsilon\n",
"\n",
"def create_sampler():\n",
" '''\n",
" Creates a sampling layer.\n",
" '''\n",
" return layers.Lambda(sample, name='sampler')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now we can compose the variational auto-encoder"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"WARNING:tensorflow:From /Users/sarahluw/venv/lib/python3.7/site-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
"Instructions for updating:\n",
"Colocations handled automatically by placer.\n"
]
}
],
"source": [
"encoder = create_encoder()\n",
"decoder = create_decoder()\n",
"sampler = create_sampler()\n",
"\n",
"x = layers.Input(shape=image_shape)\n",
"t_mean, t_log_var = encoder(x)\n",
"t = sampler([t_mean, t_log_var])\n",
"t_decoded = decoder(t)\n",
"\n",
"vae = Model(x, t_decoded, name='vae')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"and define the optimization objective `neg_variational_lower_bound`. We use the negative variational lower bound as Keras expects a loss function to be minimized."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"def neg_variational_lower_bound(x, t_decoded):\n",
" '''\n",
" Negative variational lower bound used as loss function\n",
" for training the variational auto-encoder.\n",
" \n",
" Args:\n",
" x: input images\n",
" t_decoded: reconstructed images\n",
" '''\n",
" # Reconstruction loss\n",
" rc_loss = K.sum(K.binary_crossentropy(\n",
" K.batch_flatten(x), \n",
" K.batch_flatten(t_decoded)), axis=-1)\n",
"\n",
" # Regularization term (KL divergence)\n",
" kl_loss = -0.5 * K.sum(1 + t_log_var \\\n",
" - K.square(t_mean) \\\n",
" - K.exp(t_log_var), axis=-1)\n",
" \n",
" # Average over mini-batch\n",
" return K.mean(rc_loss + kl_loss)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The variational auto-encoder is now ready to be trained."
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"colab": {
"autoexec": {
"startup": false,
"wait_interval": 0
},
"base_uri": "https://localhost:8080/",
"height": 901,
"output_extras": [
{
"item_id": 44
},
{
"item_id": 45
}
]
},
"colab_type": "code",
"executionInfo": {
"elapsed": 469640,
"status": "ok",
"timestamp": 1522825755182,
"user": {
"displayName": "Martin Krasser",
"photoUrl": "//lh6.googleusercontent.com/-zVZZRiAWOs4/AAAAAAAAAAI/AAAAAAAAAlk/Q2XGRf45rYM/s50-c-k-no/photo.jpg",
"userId": "115420131270379583938"
},
"user_tz": -120
},
"id": "IcArhIgEbNoU",
"outputId": "7fdb64c3-6c04-4b6d-bc88-32a66c86e467"
},
"outputs": [],
"source": [
"# MNIST training and validation data\n",
"(x_train, _), (x_test, _) = mnist.load_data()\n",
"\n",
"x_train = x_train.astype('float32') / 255.\n",
"x_train = x_train.reshape(x_train.shape + (1,))\n",
"\n",
"x_test = x_test.astype('float32') / 255.\n",
"x_test = x_test.reshape(x_test.shape + (1,))\n",
"\n",
"# Compile variational auto-encoder model\n",
"vae.compile(optimizer='rmsprop', loss=neg_variational_lower_bound)\n",
"\n",
"# Train variational auto-encoder with MNIST images\n",
"vae.fit(x=x_train, \n",
" y=x_train,\n",
" epochs=25,\n",
" shuffle=True,\n",
" batch_size=batch_size,\n",
" validation_data=(x_test, x_test), verbose=2)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Because we've chosen a 2-dimensional latent space, it is easy to visualize. The following plot shows the distribution of the validation set in latent space, colored by target values 0-9, the values of the digits on the validation images."
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11f86bb00>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Generate latent vectors of validation set\n",
"t_test = encoder.predict(x_test)[0]\n",
"\n",
"# Plot latent vectors colored by the value of the digits on input images\n",
"plt.scatter(t_test[:, 0], t_test[:, 1], marker='x', s=0.2, c=y_test)\n",
"plt.colorbar();"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"The latent space is organized by structural similarity of the digits i.e. structurally similar digits have a smaller distance in latent space than structurally dissimilar digits. For example, digits 4 and 9 usually differ only by a horizontal bar or curve at the top of the image.\n",
"\n",
"We can also generate new images by sampling from latent space and display a 2-dimensional manifold of digits. In the following figure, samples are drawn from the 90% confidence interval of the Gaussian prior with sampling density proportial to probability density. Visualization code was taken from [here](https://github.com/fchollet/deep-learning-with-python-notebooks/blob/master/8.4-generating-images-with-vaes.ipynb). One can clearly see the co-location of digits 4 and 9 in the bottom-right region of the image."
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x11fe30c18>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from scipy.stats import norm\n",
"\n",
"# Number of samples per dimension\n",
"n = 15 \n",
"\n",
"# Sample within 90% confidence interval of the Gaussian prior\n",
"# with sampling density proportional to probability density\n",
"grid_x = norm.ppf(np.linspace(0.05, 0.95, n))\n",
"grid_y = norm.ppf(np.linspace(0.05, 0.95, n))\n",
"\n",
"digit_size = 28\n",
"figure = np.zeros((digit_size * n, digit_size * n))\n",
"\n",
"for i, yi in enumerate(grid_x):\n",
" for j, xi in enumerate(grid_y):\n",
" t_sample = np.array([[xi, yi]])\n",
" t_sample = np.tile(t_sample, batch_size).reshape(batch_size, 2)\n",
" t_decoded = decoder.predict(t_sample, batch_size=batch_size)\n",
" digit = t_decoded[0].reshape(digit_size, digit_size)\n",
" figure[i * digit_size: (i + 1) * digit_size,\n",
" j * digit_size: (j + 1) * digit_size] = digit\n",
"\n",
"plt.figure(figsize=(10, 10))\n",
"plt.imshow(figure, cmap='Greys_r');"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## References\n",
"\n",
"\\[1\\] Dimitris G. Tzikass, Aristidis et. al. [The Variational Approximation for Bayesian Inference](http://www.cs.uoi.gr/~arly/papers/SPM08.pdf). \n",
"\\[2\\] Kevin P. Murphy. [Machine Learning, A Probabilistic Perspective](https://mitpress.mit.edu/books/machine-learning-0), Chapters 11 and 21. \n",
"\\[3\\] Christopher M. Bishop. [Pattern Recognition and Machine Learning](http://www.springer.com/de/book/9780387310732), Chapters 9 and 10. \n",
"\\[4\\] Diederik P Kingma, Max Welling [Auto-Encoding Variational Bayes](https://arxiv.org/abs/1312.6114). \n",
"\\[5\\] Gómez-Bombarelli et. al. [Automatic chemical design using a data-driven continuous representation of molecules](https://arxiv.org/abs/1610.02415). \n",
"\\[6\\] François Chollet. [Deep Learning with Python](https://www.manning.com/books/deep-learning-with-python). "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"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": "3.7.3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment