Autoencoding Generative Adversarial Networks

Lazarou, Conor

doi:10.48550/arxiv.2004.05472

Cited by 3 publications

(8 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the case where the two databases have equal number of columns and onto mapping where smaller set of column is always a subset of the larger column set, we demonstrate the performance of proposed algorithm on samples from following two data generating distributions. Gaussians with mean vectors randomly sampled from [10,20] separately. The true factor dimension is taken to be k = 10.…”

Section: Discussionmentioning

confidence: 99%

“…The Adversarial Generator-Encoder Networks (AGE) [33] model uses the divergence of two induced latent distributions as a kind of premade discriminator. Expanding this approach with multiple discriminators and generators/encoders such as done in [20] can stabilize the mapping of data points to a latent space and back, perhaps reducing the mode collapse and training difficulty of many GANs. Even though the above methods have the potential to generate augmenting data, they do not directly address all the use cases of a schema-matched dataset and do not exploit the relationships between features across databases.…”

Section: Related Workmentioning

confidence: 99%

“…We generate covariance matrices by WW T + D where W is a spherical Gaussian matrix of size (n, k) and D is a diagonal matrix with n integer values randomly drawn from the interval [1,20]. Here, k denotes the underlying…”

Section: A Appendixmentioning

confidence: 99%

See 2 more Smart Citations

Deep Learning to Jointly Schema Match, Impute, and Transform Databases

Tripathi¹,

Fritz²,

Abdelhack³

et al. 2022

Preprint

View full text Add to dashboard Cite

An applied problem facing all areas of data science is harmonizing data sources. Joining data from multiple origins with unmapped and only partially overlapping features is a prerequisite to developing and testing robust, generalizable algorithms, especially in health care. We approach this issue in the common but difficult case of numeric features such as nearly Gaussian and binary features, where unit changes and variable shift make simple matching of univariate summaries unsuccessful. We develop two novel procedures to address this problem. First, we demonstrate multiple methods of "fingerprinting" a feature based on its associations to other features. In the setting of even modest prior information, this allows most shared features to be accurately identified. Second, we demonstrate a deep learning algorithm for translation between databases. Unlike prior approaches, our algorithm takes advantage of discovered mappings while identifying surrogates for unshared features and learning transformations. In synthetic and real-world experiments using two electronic health record databases, our algorithms outperform existing baselines for matching variable sets, while jointly learning to impute unshared or transformed variables.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning to Jointly Schema Match, Impute, and Transform Databases

Tripathi¹,

Fritz²,

Abdelhack³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…This loss function for reconstruction is usually a function that calculates the pixel difference between input 𝑥 and output 𝑥 ̃. After training the autoencoder, it can be used for other purposes such as feature extraction [22]. Generative adversarial networks (GANs) are a class of generative and unsupervised neural networks that can approximate the distribution of data to generate a given dataset.…”

Section: Aeganmentioning

confidence: 99%

A New Scheme for Image Compression and Encryption Using ECIES, Henon Map, and AEGAN

Shariatzadeh

Eftekhari

Rostami

2023

Preprint

View full text Add to dashboard Cite

Providing security in the transmission of images and other multimedia data has become one of the most important scientific and practical issues. In this paper, a method for compressing and encryption images is proposed, which can safely transmit images in low-bandwidth data transmission channels. At first, using the autoencoding generative adversarial network (AEGAN) model, the images are mapped to a vector in the latent space with low dimensions. In the next step, the obtained vector is encrypted using public key encryption methods. In the proposed method, Henon chaotic map is used for permutation, which makes information transfer more secure. To evaluate the results of the proposed scheme, three criteria SSIM, PSNR, and execution time have been used.

show abstract

“…There are two main approaches for learning a bidirectional GAN: (1) adversarial feature learning (BiGAN) [8] or adversarially learned inference (ALI) [9]; (2) combining autoencoder training with GANs, e.g. VAE/GAN [22], AEGAN [23]. We focus on the latter since it usually gives better reconstruction quality and its training is more stable [36].…”

Section: Related Workmentioning

confidence: 99%

AE-StyleGAN: Improved Training of Style-Based Auto-Encoders

Han

Musunuri

Min

et al. 2021

Preprint

View full text Add to dashboard Cite

StyleGANs have shown impressive results on data generation and manipulation in recent years, thanks to its disentangled style latent space. A lot of efforts have been made in inverting a pretrained generator, where an encoder is trained ad hoc after the generator is trained in a two-stage fashion. In this paper, we focus on style-based generators asking a scientific question: Does forcing such a generator to reconstruct real data lead to more disentangled latent space and make the inversion process from image to latent space easy? We describe a new methodology to train a style-based autoencoder where the encoder and generator are optimized end-to-end. We show that our proposed model consistently outperforms baselines in terms of image inversion and generation quality. Supplementary, code, and pretrained models are available on the project website 1 .

show abstract

Autoencoding Generative Adversarial Networks

Cited by 3 publications

References 0 publications

Deep Learning to Jointly Schema Match, Impute, and Transform Databases

Deep Learning to Jointly Schema Match, Impute, and Transform Databases

A New Scheme for Image Compression and Encryption Using ECIES, Henon Map, and AEGAN

AE-StyleGAN: Improved Training of Style-Based Auto-Encoders

Contact Info

Product

Resources

About