Missing Features Reconstruction Using a Wasserstein Generative Adversarial Imputation Network

Friedjungová, Magda; Vašata, Daniel; Balatsko, Maksym; Jirina, Marcel

doi:10.1007/978-3-030-50423-6_17

Cited by 12 publications

(7 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…CollaGAN (Lee et al, 2019) proposes a collaborative GAN for missing data imputation but it focuses on image data. WGAIN (Friedjungová et al, 2020), CGAIN (Awan et al, 2021), PC-GAIN (Wang et al, 2021) and S-GAIN (Neves et al, 2021) extend GAIN in various ways. IFGAN (Qiu et al, 2020) conducts missing data imputation using a feature-specific GAN and MCFlow (Richardson et al, 2020) proposes a Monte Carlo flow method for data imputation but no theoretical result is provided.…”

Section: Related Workmentioning

confidence: 99%

FragmGAN: Generative Adversarial Nets for Fragmentary Data Imputation and Prediction

Fang¹,

Shenliao²

2022

Preprint

View full text Add to dashboard Cite

Modern scientific research and applications very often encounter "fragmentary data" which brings big challenges to imputation and prediction. By leveraging the structure of response patterns, we propose a unified and flexible framework based on Generative Adversarial Nets (GAN) to deal with fragmentary data imputation and label prediction at the same time. Unlike most of the other generative model based imputation methods that either have no theoretical guarantee or only consider Missing Completed At Random (MCAR), the proposed FragmGAN has theoretical guarantees for imputation with data Missing At Random (MAR) while no hint mechanism is needed. FragmGAN trains a predictor with the generator and discriminator simultaneously. This linkage mechanism shows significant advantages for predictive performances in extensive experiments.

show abstract

Section: Related Workmentioning

confidence: 99%

FragmGAN: Generative Adversarial Nets for Fragmentary Data Imputation and Prediction

Fang¹,

Shenliao²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Generally, in the AE, the latent space is determined by the distribution of the dataset. Intuitively, a sampling-based method in a latent space can be used to perform imputation of the missing element [24], [25], [26], [27]. The main concern here is that the distribution of the latent space is hardly represented as a closed form, so it is inevitable for the actual imputation approximation to utilize the statistical approaches such as using the average of latent variables.…”

Section: Related Workmentioning

confidence: 99%

Anytime 3D Object Reconstruction using Multi-modal Variational Autoencoder

Yu¹,

Oh²

2021

Preprint

View full text Add to dashboard Cite

For effective human-robot teaming, it is important for the robots to be able to share their visual perception with the human operators. In a harsh remote collaboration setting, however, it is especially challenging to transfer a large amount of sensory data over a low-bandwidth network in realtime, e.g., for the task of 3D shape reconstruction given 2D camera images. To reduce the burden of data transferring, data compression techniques such as autoencoder can be utilized to obtain and transmit the data in terms of latent variables in a compact form. However, due to the low-bandwidth limitation or communication delay, some of the dimensions of latent variables can be lost in transit, degenerating the reconstruction results. Moreover, in order to achieve faster transmission, an intentional over compression can be used where only partial elements of the latent variables are used. To handle these incomplete data cases, we propose a method for imputation of latent variables whose elements are partially lost or manually excluded. To perform imputation with only some dimensions of variables, exploiting prior information of the category-or instance-level is essential. In general, a prior distribution used in variational autoencoders is achieved from all of the training datapoints regardless of their labels. This type of flattened prior makes it difficult to perform imputation from the category-or instancelevel distributions. We overcome this limitation by exploiting a category-specific multi-modal prior distribution in the latent space. By finding a modal in a latent space according to the remaining elements of the latent variables, the missing elements can be sampled. We evaluate the proposed approach on the 3D object reconstruction from a single 2D image task and show that the proposed approach is robust against significant data losses.

show abstract

“…Here we introduce the WGAIN following [10] closely. Let us denote by X = R m,n,3 the space of all possible images of size m × n and three color channels (RGB) and let X be a random element of X whose distribution is denoted by P(X).…”

Section: Related Workmentioning

confidence: 99%

“…The aim of this work is to address image inpainting task using Wasserstein Generative Adversarial Imputation Network (WGAIN) that was recently introduced by the authors in [10] as a general imputation model. It is a generative imputation model which, for non-visual imputation tasks, performs comparatively to other state-of-the-art methods.…”

mentioning

confidence: 99%

Image Inpainting Using Wasserstein Generative Adversarial Imputation Network

Vašata,

Halama,

Friedjungová

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Image inpainting is one of the important tasks in computer vision which focuses on the reconstruction of missing regions in an image. The aim of this paper is to introduce an image inpainting model based on Wasserstein Generative Adversarial Imputation Network. The generator network of the model uses building blocks of convolutional layers with different dilation rates, together with skip connections that help the model reproduce fine details of the output. This combination yields a universal imputation model that is able to handle various scenarios of missingness with sufficient quality. To show this experimentally, the model is simultaneously trained to deal with three scenarios given by missing pixels at random, missing various smaller square regions, and one missing square placed in the center of the image. It turns out that our model achieves high-quality inpainting results on all scenarios. Performance is evaluated using peak signal-to-noise ratio and structural similarity index on two real-world benchmark datasets, CelebA faces and Paris StreetView. The results of our model are compared to biharmonic imputation and to some of the other state-of-the-art image inpainting methods.

show abstract

Missing Features Reconstruction Using a Wasserstein Generative Adversarial Imputation Network

Cited by 12 publications

References 20 publications

FragmGAN: Generative Adversarial Nets for Fragmentary Data Imputation and Prediction

FragmGAN: Generative Adversarial Nets for Fragmentary Data Imputation and Prediction

Anytime 3D Object Reconstruction using Multi-modal Variational Autoencoder

Image Inpainting Using Wasserstein Generative Adversarial Imputation Network

Contact Info

Product

Resources

About