DOE Patents title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Mimicking of corruption in images

Abstract

Methods and systems are provided to generate an uncorrupted version of an image given an observed image that is a corrupted version of the image. In some embodiments, a corruption mimicking (“CM”) system iteratively trains a corruption mimicking network (“CMN”) to generate corrupted images given modeled images, updates latent vectors based on differences between the corrupted images and observed images, and applies a generator to the latent vectors to generate modeled images. The training, updating, and applying are performed until modeled images that are input to the CMN result in corrupted images that approximate the observed images. Because the CMN is trained to mimic the corruption of the observed images, the final modeled images represented the uncorrupted version of the observed images.

Inventors:
; ; ;
Issue Date:
Research Org.:
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1840475
Patent Number(s):
11126895
Application Number:
16/840,358
Assignee:
Lawrence Livermore National Security, LLC (Livermore, CA)
Patent Classifications (CPCs):
G - PHYSICS G06 - COMPUTING G06N - COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
DOE Contract Number:  
AC52-07NA27344
Resource Type:
Patent
Resource Relation:
Patent File Date: 04/04/2020
Country of Publication:
United States
Language:
English

Citation Formats

Anirudh, Rushil, Bremer, Peer-Timo, Jayaraman Thiagarajan, Jayaraman, and Kailkhura, Bhavya. Mimicking of corruption in images. United States: N. p., 2021. Web.
Anirudh, Rushil, Bremer, Peer-Timo, Jayaraman Thiagarajan, Jayaraman, & Kailkhura, Bhavya. Mimicking of corruption in images. United States.
Anirudh, Rushil, Bremer, Peer-Timo, Jayaraman Thiagarajan, Jayaraman, and Kailkhura, Bhavya. Tue . "Mimicking of corruption in images". United States. https://www.osti.gov/servlets/purl/1840475.
@article{osti_1840475,
title = {Mimicking of corruption in images},
author = {Anirudh, Rushil and Bremer, Peer-Timo and Jayaraman Thiagarajan, Jayaraman and Kailkhura, Bhavya},
abstractNote = {Methods and systems are provided to generate an uncorrupted version of an image given an observed image that is a corrupted version of the image. In some embodiments, a corruption mimicking (“CM”) system iteratively trains a corruption mimicking network (“CMN”) to generate corrupted images given modeled images, updates latent vectors based on differences between the corrupted images and observed images, and applies a generator to the latent vectors to generate modeled images. The training, updating, and applying are performed until modeled images that are input to the CMN result in corrupted images that approximate the observed images. Because the CMN is trained to mimic the corruption of the observed images, the final modeled images represented the uncorrupted version of the observed images.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {2021},
month = {9}
}

Works referenced in this record:

Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks
conference, October 2017


Systems and Methods for Solving Inverse Problems Using a Coupled Autoencoder
patent-application, January 2019


Kernel-Predicting Convolutional Neural Networks for Denoising
patent-application, October 2018


Learning from Estimated High-Dynamic Range All Weather Lighting Parameters
patent-application, March 2021


Solving Linear Inverse Problems Using Gan Priors: An Algorithm with Provable Guarantees
conference, April 2018


Techniques for Upscaling Images Generated with Undetermined Down Scaling Kernels
patent-application, February 2021


Optimality and stability of the K-hyperline clustering algorithm
journal, July 2011


Image signal processor for processing images
patent, May 2020


Method of Image Completion
patent-application, August 2019


Evaluating Open-Universe Face Identification on the Web
conference, June 2013


Systems and Methods for Reconstructing Frames
patent-application, March 2021


Deep Learning Face Attributes in the Wild
conference, December 2015


Mixing matrix estimation using discriminative clustering for blind source separation
journal, January 2013


Image Processing Apparatus and Image Processing Method Thereof
patent-application, November 2020


Generating Multimodal Image Edits for a Digital Image
patent-application, August 2018


Deep Residual Learning for Image Recognition
conference, June 2016


Towards Evaluating the Robustness of Neural Networks
conference, May 2017


Ultra-High Compression of Images Based on Deep Learning
patent-application, October 2016


Image Understanding Using Sparse Representations
journal, April 2014


Training a Class-Conditional Generative Adversarial Network
patent-application, March 2021


Linear inverse problems in imaging
journal, July 2008


ReconNet: Non-Iterative Reconstruction of Images from Compressively Sensed Measurements
conference, June 2016


A Style-Based Generator Architecture for Generative Adversarial Networks
conference, June 2019


Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space?
conference, October 2019


Context Encoders: Feature Learning by Inpainting
conference, June 2016


Semantic Image Inpainting with Deep Generative Models
conference, July 2017


Generating User-Customized Items Using a Visually-Aware Image Generation Network
patent-application, August 2019


Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network
conference, July 2017


Image Processing Method and Device, Storage Medium and Electronic Device
patent-application, November 2020


Video Enhancement Using a Generator with Filters of Generative Adversarial Network
patent-application, November 2020


ImageNet Large Scale Visual Recognition Challenge
journal, April 2015


Generative Augmentation of Image Data
patent-application, September 2019


Neural Network System with Temporal Feedback for Denoising of Rendered Sequences
patent-application, April 2020


Denoising Monte Carlo Renderings Using Progressive Neural Networks
patent-application, October 2018


Surface Defect Identification Method and Apparatus
patent-application, November 2020


DeepFool: A Simple and Accurate Method to Fool Deep Neural Networks
conference, June 2016


Model Parameter Learning Device, Control Device, and Model Parameter Learning Method
patent-application, March 2020


Unsupervised Visual Domain Adaptation Using Subspace Alignment
conference, December 2013


Image-to-Image Translation with Conditional Adversarial Networks
conference, July 2017


One Network to Solve Them All — Solving Linear Inverse Problems Using Deep Projection Models
conference, October 2017


Image Distortion Correction Method and Image Distortion Correction Device Using the Same
patent-application, March 2015


Method and Apparatus for Applying Deep Learning Techniques in Video Coding, Restoration and Video Quality Analysis (VQA)
patent-application, April 2021