Search Results for author: Jonas Adler

Found 16 papers, 9 papers with code

On the unreasonable effectiveness of CNNs

no code implementations29 Jul 2020 Andreas Hauptmann, Jonas Adler

Deep learning methods using convolutional neural networks (CNN) have been successfully applied to virtually all imaging problems, and particularly in image reconstruction tasks with ill-posed and complicated imaging models.

Image Reconstruction

Kernel of CycleGAN as a principal homogeneous space

no code implementations ICLR 2020 Nikita Moriakov, Jonas Adler, Jonas Teuwen

It is known that the CycleGAN problem might admit multiple solutions, and our goal in this paper is to analyze the space of exact solutions and to give perturbation bounds for approximate solutions.

Image-to-Image Translation Translation

Kernel of CycleGAN as a Principle homogeneous space

no code implementations24 Jan 2020 Nikita Moriakov, Jonas Adler, Jonas Teuwen

It is known that the CycleGAN problem might admit multiple solutions, and our goal in this paper is to analyze the space of exact solutions and to give perturbation bounds for approximate solutions.

Image-to-Image Translation Translation

A unified representation network for segmentation with missing modalities

no code implementations19 Aug 2019 Kenneth Lau, Jonas Adler, Jens Sjölund

The second is the unified representation network: a network architecture that maps a variable number of input modalities into a unified representation that can be used for downstream tasks such as segmentation.

Multi-Scale Learned Iterative Reconstruction

1 code implementation1 Aug 2019 Andreas Hauptmann, Jonas Adler, Simon Arridge, Ozan Öktem

Applicability of these methods to large scale inverse problems is however limited by the available memory for training and extensive training times, the latter due to computationally expensive forward models.

Computed Tomography (CT)

Deep Bayesian Inversion

1 code implementation14 Nov 2018 Jonas Adler, Ozan Öktem

Characterizing statistical properties of solutions of inverse problems is essential for decision making.

Decision Making Image Reconstruction

Task adapted reconstruction for inverse problems

no code implementations27 Aug 2018 Jonas Adler, Sebastian Lunz, Olivier Verdier, Carola-Bibiane Schönlieb, Ozan Öktem

The paper considers the problem of performing a task defined on a model parameter that is only observed indirectly through noisy data in an ill-posed inverse problem.

Image Reconstruction

Deep Learning Framework for Digital Breast Tomosynthesis Reconstruction

no code implementations14 Aug 2018 Nikita Moriakov, Koen Michielsen, Jonas Adler, Ritse Mann, Ioannis Sechopoulos, Jonas Teuwen

In this study we propose an extension of the Learned Primal-Dual algorithm for digital breast tomosynthesis.

Data-driven nonsmooth optimization

1 code implementation2 Aug 2018 Sebastian Banert, Axel Ringh, Jonas Adler, Johan Karlsson, Ozan Öktem

In this work, we consider methods for solving large-scale optimization problems with a possibly nonsmooth objective function.

Optimization and Control 90C25 (Primary) 68T05, 47H05 (Secondary)

Banach Wasserstein GAN

2 code implementations NeurIPS 2018 Jonas Adler, Sebastian Lunz

Wasserstein Generative Adversarial Networks (WGANs) can be used to generate realistic samples from complicated image distributions.

A modified fuzzy C means algorithm for shading correction in craniofacial CBCT images

2 code implementations17 Jan 2018 Awais Ashfaq, Jonas Adler

Numerous image-domain correction algorithms have been proposed in the literature that use patient-specific planning CT images to estimate shading contributions in CBCT images.

Learning to solve inverse problems using Wasserstein loss

1 code implementation30 Oct 2017 Jonas Adler, Axel Ringh, Ozan Öktem, Johan Karlsson

We propose using the Wasserstein loss for training in inverse problems.

Model based learning for accelerated, limited-view 3D photoacoustic tomography

no code implementations31 Aug 2017 Andreas Hauptmann, Felix Lucka, Marta Betcke, Nam Huynh, Jonas Adler, Ben Cox, Paul Beard, Sebastien Ourselin, Simon Arridge

Recent advances in deep learning for tomographic reconstructions have shown great potential to create accurate and high quality images with a considerable speed-up.

Tomographic Reconstructions

Learned Primal-dual Reconstruction

3 code implementations20 Jul 2017 Jonas Adler, Ozan Öktem

We propose the Learned Primal-Dual algorithm for tomographic reconstruction.

SSIM

Solving ill-posed inverse problems using iterative deep neural networks

5 code implementations13 Apr 2017 Jonas Adler, Ozan Öktem

We propose a partially learned approach for the solution of ill posed inverse problems with not necessarily linear forward operators.

Cannot find the paper you are looking for? You can Submit a new open access paper.