VAEs are vulnerable to adversarial attacks; new methods improve their robustness.
problem Vulnerability of VAEs to adversarial attacks.
method Introducing methods for producing adversarially robust VAEs, including disentangling latent representations and applying them to hierarchical VAEs.
result Hierarchical VAEs produce high-fidelity autoencoders that are also adversarially robust.
Half-AVAE enhances VAE for underdetermined ICA with adversarial training.
problem Challenges in ICA under underdetermined conditions.
method Encoder-free VAE with adversarial networks and EE terms.
result Half-AVAE outperforms baseline models in underdetermined ICA.
Study shows how to manipulate VAEs for attacks and assess their robustness.
problem Adversarial attacks on Variational Autoencoders (VAE).
method Examine modifications to VAEs and propose metrics for robustness.
result Metrics to quantify the robustness of VAEs to adversarial attacks.
A new form of the variational autoencoder (VAE) is proposed, based on the symmetric Kullback-Leibler divergence. It is demonstrated that learning of the resulting symmetric VAE (sVAE) has close connections to previously developed adversarial-learning methods. This relationship helps unify the previously distinct techni…
CIVET method provides robustness guarantees for VAEs under adversarial attacks.
problem Certified probabilistic guarantees for VAEs in safety-critical applications.
method Bounding worst-case VAE error by error on support sets at the latent layer.
result CIVET outperforms state-of-the-art methods in robustness and standard performance.
Proposes a non-adversarial method for distribution matching.
problem Stability and optimization challenges in adversarial matching methods.
method Non-adversarial VAE-based matching method with alignment upper bounds.
result Demonstrates applicability of non-adversarial matching methods without modifying original architectures.
We explore methods of producing adversarial examples on deep generative models such as the variational autoencoder (VAE) and the VAE-GAN. Deep learning architectures are known to be vulnerable to adversarial examples, but previous work has focused on the application of adversarial examples to classification tasks. Deep…
Certifiably robust VAEs are trained with bounds on input perturbations.
problem Ensuring VAEs are robust to adversarial attacks.
method Derive bounds on minimal perturbation size, control parameters, and train VAEs to meet criteria.
result Certifiably robust VAEs are more robust to attacks than standard VAEs.
Improved IFA with Generative Adversarial Networks for high-dimensional latent variables.
problem Limited expressiveness of traditional VAEs in high-dimensional latent variable modeling.
method Introducing Adversarial Variational Bayes (AVB) and Importance-weighted Adversarial Variational Bayes (IWAVB) algorithms.
result IWAVB demonstrated superior expressiveness and higher likelihood compared to IWAE.
MAVENs combine GAN and VAE for better image generation and classification.
problem Improve image generation and classification using unsupervised learning.
method Introduce MAVENs, an ensemble of discriminators in a VAE-GAN network.
result Demonstrated competitive performance in image generation and classification tasks.
Paper introduces a method for learning interpretable disentangled representations using adversarial VAEs.
problem Learning interpretable and disentangled feature representations in medical applications.
method Adversarial Variational Autoencoder with total correlation constraint.
result The learned disentangled representation is interpretable and superior to state-of-the-art methods, showing improvements in disentanglement, clustering, and classification.
Develops a new robustness criterion for VAEs and provides theoretical guarantees.
problem Lack of formalization for robustness in VAEs.
method Introduces r-robustness criterion and derives reconstruction margins. result Derives theoretical guarantees for VAE robustness.
New molecular design model outperforms existing methods.
problem Designing valid, unique, and novel molecules.
method Adversarially Regularized Autoencoder (ARAE) combining latent variables from VAE and adversarial training from GAN.
result ARAE outperforms conventional models in validity, uniqueness, and novelty.
Deep generative models have achieved impressive success in recent years. Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs), as emerging families for generative model learning, have largely been considered as two distinct paradigms and received extensive independent studies respectively. This pa…
Paper proposes robust generative models using VAEs.
problem Lack of robustness in generative models.
method Formally defined robust lower bound, optimized during training.
result Generative models become more robust to adversarial attacks.
APo-VAE generates text in hyperbolic space for better hierarchical representation.
problem Lack of hierarchical structure in Euclidean embeddings for natural language.
method Adversarial Poincare Variational Autoencoder (APo-VAE) in hyperbolic latent space.
result APo-VAE outperforms Euclidean VAEs in capturing latent language hierarchies.
IDVAE combines VAE and GAN without explicit discriminator.
problem Combining VAE and GAN for reconstruction and generation.
method Shared encoder and discriminator with combined decoder and generator.
result IDVAE outperforms state-of-the-art hybrid approaches.
VAI learns efficient sampling for semi-supervised learning.
problem Efficiently acquiring labeled data in semi-supervised learning.
method Variational autoencoder (VAE) and adversarial network to learn latent space.
result Establishes new state of the art on various datasets.
VAE fails to encode typical samples, new method improves robustness.
problem VAE does not necessarily encode typical samples generated by its decoder.
method Alternative construction of variational approximation distribution, self-consistency approach.
result Encoders trained with self-consistency approach are robust to adversarial attacks.
This short article revisits some of the ideas introduced in arXiv:1701.07875 and arXiv:1705.07642 in a simple setup. This sheds some lights on the connexions between Variational Autoencoders (VAE), Generative Adversarial Networks (GAN) and Minimum Kantorovitch Estimators (MKE).
We combine conditional variational autoencoders (VAE) with adversarial censoring in order to learn invariant representations that are disentangled from nuisance/sensitive variations. In this method, an adversarial network attempts to recover the nuisance variable from the representation, which the VAE is trained to pre…
AVAE improves semi-supervised learning by generating exclusive latent codes.
problem Inadequate exclusive latent codes in traditional VAE for robust classification.
method AVAE++ generates exclusive latent codes using a combination of VAE and GAN.
result AVAE outperforms state-of-the-art models in semisupervised classification.
Enhances generative models stability and accuracy with BNPL, WMMD, and triple model.
problem Overfitting in GANs and noisy samples in VAEs.
method Bayesian non-parametric learning framework, integrating Wasserstein distance and maximum mean discrepancy.
result Superior performance across various generative tasks.
Improved VAE model for discrete data through augmented training and multiscale approach.
problem Limited capability of VAE in capturing field correlations in structured data.
method Augmented training with generated variants and multiscale VAE with multiple β values.
result Improved generation quality of VAE models through these methods.
VAEs and GANs use simple distributions and neural networks to implicitly approximate complex data distributions.
problem Approximating high-dimensional complex distributions explicitly is often intractable.
method VAEs and GANs use simple base distributions and neural networks to implicitly approximate complex distributions.
result Implicit approximation of complex distributions is crucial but introduces limitations, especially in VAEs with fixed Gaussian priors.
VAEs analyzed using harmonic analysis, showing how variance controls frequency content and robustness.
problem Understanding and optimizing VAEs for robustness and frequency control.
method Viewing VAE latent space as Gaussian space, deriving results on variance and frequency content, and demonstrating soft Lipschitz constraints.
result Increasing encoder variance reduces high frequency content and improves adversarial robustness.
Improved neural spike inference from calcium imaging data.
problem Neural spike inference from calcium imaging data.
method Importance weighted adversarial variational autoencoders (IWAE) with adversarial training.
result Adversarial IWAE methods outperform VAEs in inferring neural spikes.
Infer-AVAE infers missing user attributes from incomplete data using a novel adversarial approach.
problem Incomplete user attributes in social networks.
method Infer-AVAE combines MLP and GNNs with adversarial training to infer missing attributes.
result Infer-AVAE outperforms baselines by 7.0% in accuracy on real-world datasets.
This paper characterizes VAE training pathologies and their effects on tasks.
problem Characterizing VAE training pathologies and their impact on downstream tasks.
method Concretely characterizing conditions for VAE training pathologies and their connection to specific downstream tasks.
result Connects VAE training pathologies to specific downstream tasks like learning compressed and disentangled representations, adversarial robustness, and semi-supervised learning.
Paper proposes GAN architectures for generating vector sketches.
problem Lack of GAN architectures for generating vector sketches.
method Proposes SkeGAN and VASkeGAN architectures for vector sketch generation.
result Validated that proposed architectures generate visually appealing sketches.
Enhanced VQ-VAE generates high-fidelity images faster.
problem Generating high-fidelity images efficiently.
method Scaled VQ-VAE with fast autoregressive sampling.
result VQ-VAE generates samples with quality rivaling GANs.
A new hybrid VAE-GAN framework improves mode coverage and quality.
problem Mode collapse and poor sample quality in GANs and VAEs.
method Integrates a 'Best-of-Many-Samples' reconstruction cost and a stable synthetic likelihood estimate.
result Significant improvement in mode coverage and quality compared to hybrid VAE-GANs and plain GANs.
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…
Unbalanced GANs stabilize GAN training by pre-training the generator with VAE.
problem Stable training of GANs to avoid mode collapses and improve image quality.
method Pre-train GAN generator with VAE, balance generator and discriminator training, prevent discriminator's early convergence.
result Unbalanced GANs reduce mode collapses and outperform ordinary GANs in stability, convergence, and image quality.
VI approximates complex densities faster than classical methods.
problem Approximating complex probability densities.
method Optimization of a family of probability density functions using KL divergence.
result VI converges faster than Markov Chain Monte Carlo.
A new prior improves generative models' performance.
problem Mode collapse and poor evidence lower bound in GANs and VAEs.
method Tensor Ring Induced Prior (TRIP) that packs many Gaussians into a lattice.
result TRIP improves FID for GANs and ELBO for VAEs.
A new model combines VAE and GAN for better anomaly detection in imbalanced datasets.
problem Anomaly detection in imbalanced datasets, especially in medical applications.
method β-VAEGAN model combining VAE and GAN, kernelized SVM for anomaly scores, and deviation from Gaussian prior.
result Improved F1 score from 0.85 to 0.92 on MITBIH Arrhythmia Database. Proposes PI-VAE for solving SDEs with limited measurements.
problem Solving SDEs with limited measurements of system parameters.
method Physics-informed Variational Autoencoder (PI-VAE) integrating VAE and governing equations.
result Satisfactory accuracy and efficiency compared to PI-WGAN.
Generative Latent Nearest Neighbors (GLANN) generates better images without adversarial training.
problem Stability and mode collapse issues in GANs for image generation.
method Combines IMLE and GLO strengths to avoid GAN's weaknesses.
result Generative Latent Nearest Neighbors (GLANN) generates better images than GLO and IMLE.
Develops a new non-adversarial framework for better generative models.
problem Inaccurate approximation of target distribution in latent space.
method Tessellated Wasserstein Auto-Encoders (TWAE) using centroidal Voronoi tessellation (CVT) to tessellate latent space.
result Significantly enhances generative performance in terms of FID compared to existing models.
New method for bidirectional generative modeling using adversarial gradient estimation.
problem Bidirectional generative modeling with various f-divergences. method Adversarial gradient estimation for f-divergence optimization. result Similar algorithms for different f-divergences with varying scaling. New method uses VAEs to generate financial correlation matrices for credit portfolio VaR analysis.
problem Quantifying credit portfolio sensitivity to asset correlations.
method Employing Variational Autoencoders (VAEs) to generate synthetic financial correlation matrices.
result The VAE latent space captures crucial factors impacting portfolio diversification, especially in credit portfolio sensitivity to asset correlations.
Generative Adversarial Nets (GANs) and Variational Auto-Encoders (VAEs) provide impressive image generations from Gaussian white noise, but the underlying mathematics are not well understood. We compute deep convolutional network generators by inverting a fixed embedding operator. Therefore, they do not require to be o…
Building on the success of deep learning, two modern approaches to learn a probability model from the data are Generative Adversarial Networks (GANs) and Variational AutoEncoders (VAEs). VAEs consider an explicit probability model for the data and compute a generative distribution by maximizing a variational lower-boun…
Many deep learning algorithms can be easily fooled with simple adversarial examples. To address the limitations of existing defenses, we devised a probabilistic framework that can generate an exponentially large ensemble of models from a single model with just a linear cost. This framework takes advantage of neural net…
New model generates unseen attribute combinations from limited data.
problem Lack of generalization in deep generative models for unseen attribute combinations.
method Introduces multilinear latent conditioning to capture multiplicative interactions.
result Demonstrates effectiveness on MNIST, Fashion-MNIST, and CelebA datasets.
SimVAE trains interpretable VAEs using simulators.
problem Training interpretable generative models.
method Two-step process: train decoder to approximate simulator, then train encoder to invert it.
result Disentangled and interpretable latent space achieved.
We propose the Wasserstein Auto-Encoder (WAE)---a new algorithm for building a generative model of the data distribution. WAE minimizes a penalized form of the Wasserstein distance between the model distribution and the target distribution, which leads to a different regularizer than the one used by the Variational Aut…