Discrete-time diffusion-based generative models and score matching methods have shown promising results in modeling high-dimensional image data. Recently, Song et al. (2021) show that diffusion processes that transform data into noise can be reversed via learning the score function, i.e. the gradient of the log-density of the perturbed data. They propose to plug the learned score function into an inverse formula to define a generative diffusion process. Despite the empirical success, a theoretical underpinning of this procedure is still lacking. In this work, we approach the (continuous-time) generative diffusion directly and derive a variational framework for likelihood estimation, which includes continuous-time normalizing flows as a special case, and can be seen as an infinitely deep variational autoencoder. Under this framework, we show that minimizing the score-matching loss is equivalent to maximizing a lower bound of the likelihood of the plug-in reverse SDE proposed by Song et al. (2021), bridging the theoretical gap.
MPI Tübingen（图宾根马普研究所）；Microsoft Research Cambridge（微软剑桥研究院）；ETH Zürich（苏黎世联邦理工学院）
Variational Autoencoders (VAEs) are expressive latent variable models that can be used to learn complex probability distributions from training data. However, the quality of the resulting model crucially relies on the expressiveness of the inference model. We introduce Adversarial Variational Bayes (AVB), a technique for training Variational Autoencoders with arbitrarily expressive inference models. We achieve this by introducing an auxiliary discriminative network that allows to rephrase the maximum-likelihood-problem as a two-player game, hence establishing a principled connection between VA…
Binghamton University（宾汉姆顿大学）；IBM T.J. Watson Research Center（IBM T.J. Watson研究中心）
In this paper, we study deep generative models for effective unsupervised learning. We propose VGAN, which works by minimizing a variational lower bound of the negative log likelihood (NLL) of an energy based model (EBM), where the model density $p(\mathbf{x})$ is approximated by a variational distribution $q(\mathbf{x})$ that is easy to sample from. The training of VGAN takes a two step procedure: given $p(\mathbf{x})$, $q(\mathbf{x})$ is updated to maximize the lower bound; $p(\mathbf{x})$ is then updated one step with samples drawn from $q(\mathbf{x})$ to decrease the lower bound. VGAN is i…
Introduction to Stochastic Differential Equations for Generative Machine Learning: A Variational Perspective
Department of Biology, University of Copenhagen（哥本哈根大学生物系）；Department of Applied Mathematics and Computer Science, Technical University of Denmark（丹麦技术大学应用数学与计算机科学系）；Google DeepMind（谷歌DeepMind）；Technical University of Berlin (TU Berlin)（柏林工业大学）；Technical University of Munich（慕尼黑工业大学）
We propose a molecular generative model based on the conditional variational autoencoder for de novo molecular design. It is specialized to control multiple molecular properties simultaneously by imposing them on a latent space. As a proof of concept, we demonstrate that it can be used to generate drug-like molecules with five target properties. We were also able to adjust a single property without changing the others and to manipulate it beyond the range of the dataset.
We reinterpreting the variational inference in a new perspective. Via this way, we can easily prove that EM algorithm, VAE, GAN, AAE, ALI(BiGAN) are all special cases of variational inference. The proof also reveals the loss of standard GAN is incomplete and it explains why we need to train GAN cautiously. From that, we find out a regularization term to improve stability of GAN training.
Recent advances in semi-supervised learning with deep generative models have shown promise in generalizing from small labeled datasets ($\mathbf{x},\mathbf{y}$) to large unlabeled ones ($\mathbf{x}$). In the case where the codomain has known structure, a large unfeatured dataset ($\mathbf{y}$) is potentially available. We develop a parameter-efficient, deep semi-supervised generative model for the purpose of exploiting this untapped data source. Empirical results show improved performance in disentangling latent variable semantics as well as improved discriminative prediction on Martian spectr…
Auto-encoding generative adversarial networks (GANs) combine the standard GAN algorithm, which discriminates between real and model-generated data, with a reconstruction loss given by an auto-encoder. Such models aim to prevent mode collapse in the learned generative model by ensuring that it is grounded in all the available training data. In this paper, we develop a principle upon which auto-encoders can be combined with generative adversarial networks by exploiting the hierarchical structure of the generative model. The underlying principle shows that variational inference can be used a basi…
University of Amsterdam（阿姆斯特丹大学）；Linköping University（林雪平大学）
Generative flow networks (GFNs) are a class of models for sequential sampling of composite objects, which approximate a target distribution that is defined in terms of an energy function or a reward. GFNs are typically trained using a flow matching or trajectory balance objective, which matches forward and backward transition models over trajectories. In this work, we define variational objectives for GFNs in terms of the Kullback-Leibler (KL) divergences between the forward and backward distribution. We show that variational inference in GFNs is equivalent to minimizing the trajectory balance…
Generative neural samplers are probabilistic models that implement sampling using feedforward neural networks: they take a random input vector and produce a sample from a probability distribution defined by the network weights. These models are expressive and allow efficient computation of samples and derivatives, but cannot be used for computing likelihoods or for marginalization. The generative-adversarial training method allows to train such models through the use of an auxiliary discriminative neural network. We show that the generative-adversarial approach is a special case of an existing…
We develop a novel probabilistic generative model based on the variational autoencoder approach. Notable aspects of our architecture are: a novel way of specifying the latent variables prior, and the introduction of an ordinality enforcing unit. We describe how to do supervised, unsupervised and semi-supervised learning, and nominal and ordinal classification, with the model. We analyze generative properties of the approach, and the classification effectiveness under nominal and ordinal classification, using two benchmark datasets. Our results show that our model can achieve comparable results…
University of Liège（列日大学）；New York University（纽约大学）
Complex computer simulators are increasingly used across fields of science as generative models tying parameters of an underlying theory to experimental observations. Inference in this setup is often difficult, as simulators rarely admit a tractable density or likelihood function. We introduce Adversarial Variational Optimization (AVO), a likelihood-free inference algorithm for fitting a non-differentiable generative model incorporating ideas from generative adversarial networks, variational optimization and empirical Bayes. We adapt the training procedure of generative adversarial networks by…
We propose Unbalanced GANs, which pre-trains the generator of the generative adversarial network (GAN) using variational autoencoder (VAE). We guarantee the stable training of the generator by preventing the faster convergence of the discriminator at early epochs. Furthermore, we balance between the generator and the discriminator at early epochs and thus maintain the stabilized training of GANs. We apply Unbalanced GANs to well known public datasets and find that Unbalanced GANs reduce mode collapses. We also show that Unbalanced GANs outperform ordinary GANs in terms of stabilized learning, …
University of Applied Sciences and Arts Western Switzerland (HES-SO)（瑞士西部应用科学与艺术大学（HES-SO））
Integrating physics models within machine learning models holds considerable promise toward learning robust models with improved interpretability and abilities to extrapolate. In this work, we focus on the integration of incomplete physics models into deep generative models. In particular, we introduce an architecture of variational autoencoders (VAEs) in which a part of the latent space is grounded by physics. A key technical challenge is to strike a balance between the incomplete physics and trainable components such as neural networks for ensuring that the physics part is used in a meaningf…
Training and using modern neural-network based latent-variable generative models (like Variational Autoencoders) often require simultaneously training a generative direction along with an inferential(encoding) direction, which approximates the posterior distribution over the latent variables. Thus, the question arises: how complex does the inferential model need to be, in order to be able to accurately model the posterior distribution of a given generative model? In this paper, we identify an important property of the generative map impacting the required size of the encoder. We show that if t…
ETH Zurich（苏黎世联邦理工学院）；University of Zurich（苏黎世大学）；Stanford University（斯坦福大学）；Stanford Stroke Center（斯坦福中风中心）；Institute for Biomedical Engineering（生物医学工程研究所）
We show that high quality, diverse and realistic-looking diffusion-weighted magnetic resonance images can be synthesized using deep generative models. Based on professional neuroradiologists' evaluations and diverse metrics with respect to quality and diversity of the generated synthetic brain images, we present two networks, the Introspective Variational Autoencoder and the Style-Based GAN, that qualify for data augmentation in the medical field, where information is saved in a dispatched and inhomogeneous way and access to it is in many aspects restricted.
With the increasingly widespread deployment of generative models, there is a mounting need for a deeper understanding of their behaviors and limitations. In this paper, we expose the limitations of Variational Autoencoders (VAEs), which consistently fail to learn marginal distributions in both latent and visible spaces. We show this to be a consequence of learning by matching conditional distributions, and the limitations of explicit model and posterior distributions. It is popular to consider Generative Adversarial Networks (GANs) as a means of overcoming these limitations, leading to hybrids…
Université Paris Dauphine - PSL Research University（巴黎第九大学-巴黎文理研究大学）；CEREMADE（CEREMADE（巴黎第九大学应用数学研究中心））
The quality of generative models (such as Generative adversarial networks and Variational Auto-Encoders) depends heavily on the choice of a good probability distance. However some popular metrics like the Wasserstein or the Sliced Wasserstein distances, the Jensen-Shannon divergence, the Kullback-Leibler divergence, lack convenient properties such as (geodesic) convexity, fast evaluation and so on. To address these shortcomings, we introduce a class of distances that have built-in convexity. We investigate the relationship with some known paradigms (sliced distances - a synonym for Radon dista…
University of Manitoba（曼尼托巴大学）；TRIUMF；University of British Columbia（不列颠哥伦比亚大学）
Matter-antimatter asymmetry is one of the major unsolved problems in physics that can be probed through precision measurements of charge-parity symmetry violation at current and next-generation neutrino oscillation experiments. In this work, we demonstrate the capability of variational autoencoders and normalizing flows to approximate the generative distribution of simulated data for water Cherenkov detectors commonly used in these experiments. We study the performance of these methods and their applicability for semi-supervised learning and synthetic data generation.
The ability to extract generative parameters from high-dimensional fields of data in an unsupervised manner is a highly desirable yet unrealized goal in computational physics. This work explores the use of variational autoencoders (VAEs) for non-linear dimension reduction with the specific aim of {\em disentangling} the low-dimensional latent variables to identify independent physical parameters that generated the data. A disentangled decomposition is interpretable, and can be transferred to a variety of tasks including generative modeling, design optimization, and probabilistic reduced order …
Generative adversarial networks (GANs) have given us a great tool to fit implicit generative models to data. Implicit distributions are ones we can sample from easily, and take derivatives of samples with respect to model parameters. These models are highly expressive and we argue they can prove just as useful for variational inference (VI) as they are for generative modelling. Several papers have proposed GAN-like algorithms for inference, however, connections to the theory of VI are not always well understood. This paper provides a unifying review of existing algorithms establishing connecti…
Generative Modeling of Quantum Distribution with Functional Flow Matching
Graduate School of Data Science, Seoul National University, Seoul, Korea（首尔国立大学数据科学研究生院）；Department of Statistics and Data Science, Yonsei University, Seoul, Korea（延世大学统计与数据科学系）；Department of Applied Statistics, Yonsei University, Seoul, Korea（延世大学应用统计系）；Google Research, Mountain View, California, United States（谷歌研究）
Harvard University（哈佛大学）；Massachusetts Institute of Technology（麻省理工学院）
Amortized variational inference (AVI) replaces instance-specific local inference with a global inference network. While AVI has enabled efficient training of deep generative models such as variational autoencoders (VAE), recent empirical work suggests that inference networks can produce suboptimal variational parameters. We propose a hybrid approach, to use AVI to initialize the variational parameters and run stochastic variational inference (SVI) to refine them. Crucially, the local SVI procedure is itself differentiable, so the inference network and generative model can be trained end-to-end…
Massachusetts Institute of Technology（麻省理工学院）；MIT-IBM Watson AI Lab（麻省理工学院-IBM沃森人工智能实验室）
Hierarchical Bayesian methods can unify many related tasks (e.g. k-shot classification, conditional and unconditional generation) as inference within a single generative model. However, when this generative model is expressed as a powerful neural network such as a PixelCNN, we show that existing learning techniques typically fail to effectively use latent variables. To address this, we develop a modification of the Variational Autoencoder in which encoded observations are decoded to new elements from the same class. This technique, which we call a Variational Homoencoder (VHE), produces a hier…
Mila & DIRO, University of Montreal（蒙特利尔大学 Mila 与 DIRO）；Facebook Artificial Intelligence Research（Facebook 人工智能研究院）；Canada CIFAR AI Chair（加拿大 CIFAR 人工智能教席）
Generative adversarial networks (GANs) form a generative modeling approach known for producing appealing samples, but they are notably difficult to train. One common way to tackle this issue has been to propose new formulations of the GAN objective. Yet, surprisingly few studies have looked at optimization methods designed for this adversarial training. In this work, we cast GAN optimization problems in the general variational inequality framework. Tapping into the mathematical programming literature, we counter some common misconceptions about the difficulties of saddle point optimization and…
Technical University of Munich（慕尼黑工业大学）；Mercateo AG（Mercateo公司）
Variational autoencoder (VAE) is a widely used generative model for learning latent representations. Burda et al. in their seminal paper showed that learning capacity of VAE is limited by over-pruning. It is a phenomenon where a significant number of latent variables fail to capture any information about the input data and the corresponding hidden units become inactive. This adversely affects learning diverse and interpretable latent representations. As variational graph autoencoder (VGAE) extends VAE for graph-structured data, it inherits the over-pruning problem. In this paper, we adopt a mo…
University of Toronto（多伦多大学）；Canadian Institute for Advanced Research（加拿大高等研究院）
We consider the problem of learning deep generative models from data. We formulate a method that generates an independent sample via a single feedforward pass through a multilayer perceptron, as in the recently proposed generative adversarial networks (Goodfellow et al., 2014). Training a generative adversarial network, however, requires careful optimization of a difficult minimax program. Instead, we utilize a technique from statistical hypothesis testing known as maximum mean discrepancy (MMD), which leads to a simple objective that can be interpreted as matching all orders of statistics bet…
Microsoft Search, Assistance and Intelligence（微软搜索、辅助与智能部门（Microsoft Search, Assistance and Intelligence））
We consider the problem of diversifying automated reply suggestions for a commercial instant-messaging (IM) system (Skype). Our conversation model is a standard matching based information retrieval architecture, which consists of two parallel encoders to project messages and replies into a common feature representation. During inference, we select replies from a fixed response set using nearest neighbors in the feature space. To diversify responses, we formulate the model as a generative latent variable model with Conditional Variational Auto-Encoder (M-CVAE). We propose a constrained-sampling…
Variational Rank Reduction Autoencoders for Generative Thermal Design
ESI Group-UZ Chair of the National Strategy on Artificial Intelligence（ESI集团-乌兹主席国家人工智能战略办公室）；Aragon Institute of Engineering Research (I3A)（阿拉贡工程研究所（I3A））；Universidad de Zaragoza（萨拉戈萨大学）；PIMM Lab（PIMM实验室）；ENSAM Institute of Technology（ENSAM技术研究所）
A Bayesian Non-parametric Approach to Generative Models: Integrating Variational Autoencoder and Generative Adversarial Networks using Wasserstein and Maximum Mean Discrepancy
Department of Statistics and Actuarial Science University of Hong Kong（统计与精算科学系香港大学）
The Variational AutoEncoder (VAE) learns simultaneously an inference and a generative model, but only one of these models can be learned at optimum, this behaviour is associated to the ELBO learning objective, that is optimised by a non-informative generator. In order to solve such an issue, we provide a learning objective, learning a maximal informative generator while maintaining bounded the network capacity: the Variational InfoMax (VIM). The contribution of the VIM derivation is twofold: an objective learning both an optimal inference and generative model and the explicit definition of the…
The Variational Autoencoder (VAE) is a powerful framework for learning probabilistic latent variable generative models. However, typical assumptions on the approximate posterior distribution of the encoder and/or the prior, seriously restrict its capacity for inference and generative modeling. Variational inference based on neural autoregressive models respects the conditional dependencies of the exact posterior, but this flexibility comes at a cost: such models are expensive to train in high-dimensional regimes and can be slow to produce samples. In this work, we introduce an orthogonal solut…
University of British Columbia（不列颠哥伦比亚大学）；MIT（麻省理工学院）
We introduce the thermodynamic variational objective (TVO) for learning in both continuous and discrete deep generative models. The TVO arises from a key connection between variational inference and thermodynamic integration that results in a tighter lower bound to the log marginal likelihood than the standard variational variational evidence lower bound (ELBO) while remaining as broadly applicable. We provide a computationally efficient gradient estimator for the TVO that applies to continuous, discrete, and non-reparameterizable distributions and show that the objective functions used in var…
Gatsby Computational Neuroscience Unit（盖茨比计算神经科学部）；University College London（伦敦大学学院）；Amazon Web Services（亚马逊云科技）
Score matching (SM) is a convenient method for training flexible probabilistic models, which is often preferred over the traditional maximum-likelihood (ML) approach. However, these models are less interpretable than normalized models; as such, training robustness is in general difficult to assess. We present a critical study of existing variational SM objectives, showing catastrophic failure on a wide range of datasets and network architectures. Our theoretical insights on the objectives emerge directly from their equivalent autoencoding losses when optimizing variational autoencoder (VAE) mo…
