Search Results for author: Carola-Bibiane Schönlieb

Found 195 papers, 62 papers with code

Data-driven approaches to inverse problems

no code implementations13 Jun 2025 Carola-Bibiane Schönlieb, Zakhar Shumaylov

The first part of these notes will provide an introduction to inverse problems, discuss classical solution strategies, and present some applications.

Computational Efficiency

Conservation-preserved Fourier Neural Operator through Adaptive Correction

no code implementations30 May 2025 Chaoyu Liu, Yangming Li, Zhongying Deng, Chris Budd, Carola-Bibiane Schönlieb

It ensures that the outputs exactly satisfy the goal conservation law and allow for more flexibility and adaptivity for the model to correct the outputs.

Smooth Model Compression without Fine-Tuning

no code implementations30 May 2025 Christina Runkel, Natacha Kuete Meli, Jovita Lukasik, Ander Biguri, Carola-Bibiane Schönlieb, Michael Moeller

Compressing and pruning large machine learning models has become a critical step towards their deployment in real-world applications.

model Model Compression

Deep Spectral Prior

no code implementations26 May 2025 Yanqi Cheng, Tieyong Zeng, Pietro Lio, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

We introduce Deep Spectral Prior (DSP), a new formulation of Deep Image Prior (DIP) that redefines image reconstruction as a frequency-domain alignment problem.

Denoising Image Reconstruction +2

Multi-Level Monte Carlo Training of Neural Operators

no code implementations19 May 2025 James Rowbottom, Stefania Fresca, Pietro Lio, Carola-Bibiane Schönlieb, Nicolas Boullé

Our numerical experiments on a range of state-of-the-art models and test-cases demonstrate improved computational efficiency compared to traditional single-resolution training approaches, and highlight the existence of a Pareto curve between accuracy and computational time, related to the number of samples per resolution.

Computational Efficiency Operator learning

Approximation theory for 1-Lipschitz ResNets

no code implementations17 May 2025 Davide Murari, Takashi Furuya, Carola-Bibiane Schönlieb

In this paper, we focus on 1-Lipschitz residual networks (ResNets) based on explicit Euler steps of negative gradient flows and study their approximation capabilities.

Potential Contrast: Properties, Equivalences, and Generalization to Multiple Classes

1 code implementation2 May 2025 Wallace Peaslee, Anna Breger, Carola-Bibiane Schönlieb

Potential contrast is typically used as an image quality measure and quantifies the maximal possible contrast between samples from two classes of pixels in an image after an arbitrary grayscale transformation.

Brain Foundation Models with Hypergraph Dynamic Adapter for Brain Disease Analysis

no code implementations1 May 2025 Zhongying Deng, Haoyu Wang, Ziyan Huang, Lipei Zhang, Angelica I. Aviles-Rivero, Chaoyu Liu, Junjun He, Zoe Kourtzi, Carola-Bibiane Schönlieb

Brain diseases, such as Alzheimer's disease and brain tumors, present profound challenges due to their complexity and societal impact.

On Oversquashing in Graph Neural Networks Through the Lens of Dynamical Systems

1 code implementation AAAI 2025 Alessio Gravina, Moshe Eliasof, Claudio Gallicchio, Davide Bacciu, Carola-Bibiane Schönlieb

A common problem in Message-Passing Neural Networks is oversquashing -- the limited ability to facilitate effective information flow between distant nodes.

SurvSurf: a partially monotonic neural network for first-hitting time prediction of intermittently observed discrete and continuous sequential events

no code implementations7 Apr 2025 Yichen Kelly Chen, Sören Dittmer, Kinga Bernatowicz, Josep Arús-Pous, Kamen Bliznashki, John Aston, James H. F. Rudd, Carola-Bibiane Schönlieb, James Jones, Michael Roberts

We propose a neural-network based survival model (SurvSurf) specifically designed for direct and simultaneous probabilistic prediction of the first hitting time of sequential events from baseline.

D2SA: Dual-Stage Distribution and Slice Adaptation for Efficient Test-Time Adaptation in MRI Reconstruction

no code implementations25 Mar 2025 Lipei Zhang, Rui Sun, Zhongying Deng, Yanqi Cheng, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

Variations in Magnetic resonance imaging (MRI) scanners and acquisition protocols cause distribution shifts that degrade reconstruction performance on unseen data.

MRI Reconstruction Self-Supervised Learning +1

Towards Efficient Training of Graph Neural Networks: A Multiscale Approach

no code implementations25 Mar 2025 Eshed Gal, Moshe Eliasof, Carola-Bibiane Schönlieb, Eldad Haber, Eran Treister

Graph Neural Networks (GNNs) have emerged as a powerful tool for learning and inferring from graph-structured data, and are widely used in a variety of applications, often considering large amounts of data and large graphs.

Enhancing Fourier Neural Operators with Local Spatial Features

1 code implementation22 Mar 2025 Chaoyu Liu, Davide Murari, Lihao Liu, Yangming Li, Chris Budd, Carola-Bibiane Schönlieb

To address this limitation, we introduce a convolutional neural network (CNN)-based feature pre-extractor to capture LSFs directly from input data, resulting in a hybrid architecture termed \textit{Conv-FNO}.

Computational Efficiency

Enhanced Denoising and Convergent Regularisation Using Tweedie Scaling

no code implementations7 Mar 2025 Naïl Khelifa, Ferdia Sherry, Carola-Bibiane Schönlieb

The inherent ill-posed nature of image reconstruction problems, due to limitations in the physical acquisition process, is typically addressed by introducing a regularisation term that incorporates prior knowledge about the underlying image.

Denoising Image Reconstruction

Implicit U-KAN2.0: Dynamic, Efficient and Interpretable Medical Image Segmentation

no code implementations5 Mar 2025 Chun-Wun Cheng, Yining Zhao, Yanqi Cheng, Javier Montoya, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

In the SONO-MultiKAN phase, we integrate the second-order NODEs and MultiKAN layer as the core computational block to enhance interpretability and representation power.

Decoder Image Segmentation +2

Generalized Lie Symmetries in Physics-Informed Neural Operators

1 code implementation1 Feb 2025 Amy Xiang Wang, Zakhar Shumaylov, Peter Zaika, Ferdia Sherry, Carola-Bibiane Schönlieb

Physics-informed neural operators (PINOs) have emerged as powerful tools for learning solution operators of partial differential equations (PDEs).

Generative Unordered Flow for Set-Structured Data Generation

no code implementations29 Jan 2025 Yangming Li, Carola-Bibiane Schönlieb

Flow-based generative models have demonstrated promising performance across a broad spectrum of data modalities (e. g., image and text).

Inverse Evolution Data Augmentation for Neural PDE Solvers

no code implementations24 Jan 2025 Chaoyu Liu, Chris Budd, Carola-Bibiane Schönlieb

In this paper, we propose a novel data augmentation method specifically designed for training neural operators on evolution equations.

Data Augmentation

Review and Recommendations for using Artificial Intelligence in Intracoronary Optical Coherence Tomography Analysis

no code implementations24 Jan 2025 Xu Chen, Yuan Huang, Benn Jessney, Jason Sangha, Sophie Gu, Carola-Bibiane Schönlieb, Martin Bennett, Michael Roberts

Artificial intelligence (AI) methodologies hold great promise for the rapid and accurate diagnosis of coronary artery disease (CAD) from intravascular optical coherent tomography (IVOCT) images.

Diagnostic

GRAMA: Adaptive Graph Autoregressive Moving Average Models

no code implementations22 Jan 2025 Moshe Eliasof, Alessio Gravina, Andrea Ceni, Claudio Gallicchio, Davide Bacciu, Carola-Bibiane Schönlieb

Graph State Space Models (SSMs) have recently been introduced to enhance Graph Neural Networks (GNNs) in modeling long-range interactions.

State Space Models

Towards Foundation Models on Graphs: An Analysis on Cross-Dataset Transfer of Pretrained GNNs

no code implementations23 Dec 2024 Fabrizio Frasca, Fabian Jogl, Moshe Eliasof, Matan Ostrovsky, Carola-Bibiane Schönlieb, Thomas Gärtner, Haggai Maron

To develop a preliminary understanding towards Graph Foundation Models, we study the extent to which pretrained Graph Neural Networks can be applied across datasets, an effort requiring to be agnostic to dataset-specific features and their encodings.

Symplectic Neural Flows for Modeling and Discovery

no code implementations21 Dec 2024 Priscilla Canizares, Davide Murari, Carola-Bibiane Schönlieb, Ferdia Sherry, Zakhar Shumaylov

SympFlow allows for two key applications: (i) providing a time-continuous symplectic approximation of the exact flow of a Hamiltonian system--purely based on the differential equations it satisfies, and (ii) approximating the flow map of an unknown Hamiltonian system relying on trajectory data.

Cross-Modal Few-Shot Learning with Second-Order Neural Ordinary Differential Equations

no code implementations20 Dec 2024 Yi Zhang, Chun-Wun Cheng, Junyi He, Zhihai He, Carola-Bibiane Schönlieb, Yuyan Chen, Angelica I Aviles-Rivero

We introduce SONO, a novel method leveraging Second-Order Neural Ordinary Differential Equations (Second-Order NODEs) to enhance cross-modal few-shot learning.

Few-Shot Learning Image Augmentation

Benchmarking learned algorithms for computed tomography image reconstruction tasks

no code implementations11 Dec 2024 Maximilian B. Kiss, Ander Biguri, Zakhar Shumaylov, Ferdia Sherry, K. Joost Batenburg, Carola-Bibiane Schönlieb, Felix Lucka

With this benchmarking study, we provide an evaluation of a range of algorithms representative for different categories of learned reconstruction methods on a recently published dataset of real-world experimental CT measurements.

Benchmarking Computed Tomography (CT) +3

Training Data Reconstruction: Privacy due to Uncertainty?

no code implementations11 Dec 2024 Christina Runkel, Kanchana Vaishnavi Gandikota, Jonas Geiping, Carola-Bibiane Schönlieb, Michael Moeller

We demonstrate that our formulation as well as previous approaches highly depend on the initialisation of the training images $x$ to reconstruct.

valid

Towards Graph Foundation Models: A Study on the Generalization of Positional and Structural Encodings

no code implementations10 Dec 2024 Billy Joe Franks, Moshe Eliasof, Semih Cantürk, Guy Wolf, Carola-Bibiane Schönlieb, Sophie Fellenz, Marius Kloft

Recent advances in integrating positional and structural encodings (PSEs) into graph neural networks (GNNs) have significantly enhanced their performance across various graph learning tasks.

Benchmarking Graph Learning

You KAN Do It in a Single Shot: Plug-and-Play Methods with Single-Instance Priors

no code implementations9 Dec 2024 Yanqi Cheng, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

The use of Plug-and-Play (PnP) methods has become a central approach for solving inverse problems, with denoisers serving as regularising priors that guide optimisation towards a clean solution.

Denoising Kolmogorov-Arnold Networks +1

On the Utilization of Unique Node Identifiers in Graph Neural Networks

no code implementations4 Nov 2024 Maya Bechler-Speicher, Moshe Eliasof, Carola-Bibiane Schönlieb, Ran Gilad-Bachrach, Amir Globerson

Graph Neural Networks have inherent representational limitations due to their message-passing structure.

Hamiltonian Matching for Symplectic Neural Integrators

no code implementations23 Oct 2024 Priscilla Canizares, Davide Murari, Carola-Bibiane Schönlieb, Ferdia Sherry, Zakhar Shumaylov

Hamilton's equations of motion form a fundamental framework in various branches of physics, including astronomy, quantum mechanics, particle physics, and climate science.

Astronomy

Pullback Flow Matching on Data Manifolds

no code implementations6 Oct 2024 Friso de Kruiff, Erik Bekkers, Ozan Öktem, Carola-Bibiane Schönlieb, Willem Diepeveen

We propose Pullback Flow Matching (PFM), a novel framework for generative modeling on data manifolds.

Drug Discovery

Lie Algebra Canonicalization: Equivariant Neural Operators under arbitrary Lie Groups

no code implementations3 Oct 2024 Zakhar Shumaylov, Peter Zaika, James Rowbottom, Ferdia Sherry, Melanie Weber, Carola-Bibiane Schönlieb

In the context of PDE solvers, recent works have shown that Lie point symmetries can be a useful inductive bias for Physics-Informed Neural Networks (PINNs) through data and loss augmentation.

image-classification Image Classification +1

Mamba Neural Operator: Who Wins? Transformers vs. State-Space Models for PDEs

no code implementations3 Oct 2024 Chun-Wun Cheng, Jiahao Huang, Yi Zhang, Guang Yang, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

Partial differential equations (PDEs) are widely used to model complex physical systems, but solving them efficiently remains a significant challenge.

Mamba State Space Models

Score-based Pullback Riemannian Geometry: Extracting the Data Manifold Geometry using Anisotropic Flows

no code implementations2 Oct 2024 Willem Diepeveen, Georgios Batzolis, Zakhar Shumaylov, Carola-Bibiane Schönlieb

Through numerical experiments on diverse datasets, including image data, we demonstrate that the proposed framework produces high-quality geodesics passing through the data support, reliably estimates the intrinsic dimension of the data manifold, and provides a global chart of the manifold.

Representation Learning

Learning Regularization for Graph Inverse Problems

no code implementations19 Aug 2024 Moshe Eliasof, Md Shahriar Rahim Siddiqui, Carola-Bibiane Schönlieb, Eldad Haber

In recent years, Graph Neural Networks (GNNs) have been utilized for various applications ranging from drug discovery to network design and social networks.

Drug Discovery

Deep Generative Classification of Blood Cell Morphology

2 code implementations16 Aug 2024 Simon Deltadahl, Julian Gilbey, Christine Van Laer, Nancy Boeckx, Mathie Leers, Tanya Freeman, Laura Aiken, Timothy Farren, Matthew Smith, Mohamad Zeina, BloodCounts consortium, James HF Rudd, Concetta Piazzese, Joseph Taylor, Nicholas Gleadall, Carola-Bibiane Schönlieb, Suthesh Sivapalaratnam, Michael Roberts, Parashkev Nachev

Accurate classification of haematological cells is critical for diagnosing blood disorders, but presents significant challenges for machine automation owing to the complexity of cell morphology, heterogeneities of biological, pathological, and imaging characteristics, and the imbalance of cell type frequencies.

Anomaly Detection Classification +4

Learned denoising with simulated and experimental low-dose CT data

no code implementations15 Aug 2024 Maximilian B. Kiss, Ander Biguri, Carola-Bibiane Schönlieb, K. Joost Batenburg, Felix Lucka

The study furthermore suggests the need for more sophisticated noise simulation approaches to bridge the gap between simulated and real-world data in CT image denoising applications and gives insights into the challenges and opportunities in leveraging simulated data for machine learning in computational imaging.

Computed Tomography (CT) Image Denoising

Blessing of Dimensionality for Approximating Sobolev Classes on Manifolds

no code implementations13 Aug 2024 Hong Ye Tan, Subhadip Mukherjee, Junqi Tang, Carola-Bibiane Schönlieb

The manifold hypothesis says that natural high-dimensional data lie on or around a low-dimensional manifold.

Contrastive Learning with Adaptive Neighborhoods for Brain Age Prediction on 3D Stiffness Maps

no code implementations1 Aug 2024 Jakob Träuble, Lucy Hiscox, Curtis Johnson, Carola-Bibiane Schönlieb, Gabriele Kaminski Schierle, Angelica Aviles-Rivero

In the field of neuroimaging, accurate brain age prediction is pivotal for uncovering the complexities of brain aging and pinpointing early indicators of neurodegenerative conditions.

Contrastive Learning Self-Supervised Learning

NODE-Adapter: Neural Ordinary Differential Equations for Better Vision-Language Reasoning

no code implementations11 Jul 2024 Yi Zhang, Chun-Wun Cheng, Ke Yu, Zhihai He, Carola-Bibiane Schönlieb, Angelica I. Aviles-Rivero

To fully leverage both visual and textual modalities and estimate class prototypes more effectively and accurately, we divide our method into two stages: cross-modal prototype construction and cross-modal prototype optimization using neural ordinary differential equations.

Domain Generalization Human-Object Interaction Detection +1

DiGRAF: Diffeomorphic Graph-Adaptive Activation Function

2 code implementations2 Jul 2024 Krishna Sri Ipsit Mantri, Xinzhi Wang, Carola-Bibiane Schönlieb, Bruno Ribeiro, Beatrice Bevilacqua, Moshe Eliasof

In this paper, we propose a novel activation function tailored specifically for graph data in Graph Neural Networks (GNNs).

Computational Efficiency

Spatiotemporal Graph Neural Network Modelling Perfusion MRI

no code implementations10 Jun 2024 Ruodan Yan, Carola-Bibiane Schönlieb, Chao Li

Perfusion MRI (pMRI) offers valuable insights into tumor vascularity and promises to predict tumor genotypes, thus benefiting prognosis for glioma patients, yet effective models tailored to 4D pMRI are still lacking.

Graph Neural Network Graph structure learning +1

Optimised ProPainter for Video Diminished Reality Inpainting

no code implementations4 Jun 2024 Pengze Li, Lihao Liu, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

In this paper, part of the DREAMING Challenge - Diminished Reality for Emerging Applications in Medicine through Inpainting, we introduce a refined video inpainting technique optimised from the ProPainter method to meet the specialised demands of medical imaging, specifically in the context of oral and maxillofacial surgery.

Video Inpainting

FedMAP: Unlocking Potential in Personalized Federated Learning through Bi-Level MAP Optimization

1 code implementation29 May 2024 Fan Zhang, Carlos Esteve-Yagüe, Sören Dittmer, Carola-Bibiane Schönlieb, Michael Roberts

This study contributes to PFL by establishing a solid theoretical foundation for the proposed method and offering a robust, ready-to-use framework that effectively addresses the challenges posed by non-IID data in FL.

Personalized Federated Learning

When AI Eats Itself: On the Caveats of AI Autophagy

no code implementations15 May 2024 Xiaodan Xing, Fadong Shi, Jiahao Huang, Yinzhe Wu, Yang Nan, Sheng Zhang, Yingying Fang, Mike Roberts, Carola-Bibiane Schönlieb, Javier Del Ser, Guang Yang

Generative Artificial Intelligence (AI) technologies and large models are producing realistic outputs across various domains, such as images, text, speech, and music.

Continuous Learned Primal Dual

no code implementations3 May 2024 Christina Runkel, Ander Biguri, Carola-Bibiane Schönlieb

Neural ordinary differential equations (Neural ODEs) propose the idea that a sequence of layers in a neural network is just a discretisation of an ODE, and thus can instead be directly modelled by a parameterised ODE.

Computed Tomography (CT) CT Reconstruction

Tackling Graph Oversquashing by Global and Local Non-Dissipativity

no code implementations2 May 2024 Alessio Gravina, Moshe Eliasof, Claudio Gallicchio, Davide Bacciu, Carola-Bibiane Schönlieb

A common problem in Message-Passing Neural Networks is oversquashing -- the limited ability to facilitate effective information flow between distant nodes.

GRANOLA: Adaptive Normalization for Graph Neural Networks

no code implementations20 Apr 2024 Moshe Eliasof, Beatrice Bevilacqua, Carola-Bibiane Schönlieb, Haggai Maron

In recent years, significant efforts have been made to refine the design of Graph Neural Network (GNN) layers, aiming to overcome diverse challenges, such as limited expressive power and oversmoothing.

Graph Neural Network

Unsupervised Training of Convex Regularizers using Maximum Likelihood Estimation

no code implementations8 Apr 2024 Hong Ye Tan, Ziruo Cai, Marcelo Pereyra, Subhadip Mukherjee, Junqi Tang, Carola-Bibiane Schönlieb

However, many such methods require the availability of ground truth data, which may be unavailable or expensive, leading to a fundamental barrier that can not be bypassed by choice of architecture.

Denoising

Bilevel Hypergraph Networks for Multi-Modal Alzheimer's Diagnosis

no code implementations19 Mar 2024 Angelica I. Aviles-Rivero, Chun-Wun Cheng, Zhongying Deng, Zoe Kourtzi, Carola-Bibiane Schönlieb

Early detection of Alzheimer's disease's precursor stages is imperative for significantly enhancing patient outcomes and quality of life.

Biophysics Informed Pathological Regularisation for Brain Tumour Segmentation

no code implementations14 Mar 2024 Lipei Zhang, Yanqi Cheng, Lihao Liu, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

Recent advances in deep learning have significantly improved brain tumour segmentation techniques; however, the results still lack confidence and robustness as they solely consider image data without biophysical priors or pathological information.

Segmentation

HAMLET: Graph Transformer Neural Operator for Partial Differential Equations

no code implementations5 Feb 2024 Andrey Bryutkin, Jiahao Huang, Zhongying Deng, Guang Yang, Carola-Bibiane Schönlieb, Angelica Aviles-Rivero

We present a novel graph transformer framework, HAMLET, designed to address the challenges in solving partial differential equations (PDEs) using neural networks.

Weakly Convex Regularisers for Inverse Problems: Convergence of Critical Points and Primal-Dual Optimisation

no code implementations1 Feb 2024 Zakhar Shumaylov, Jeremy Budd, Subhadip Mukherjee, Carola-Bibiane Schönlieb

Variational regularisation is the primary method for solving inverse problems, and recently there has been considerable work leveraging deeply learned regularisation for enhanced performance.

Computed Tomography (CT) CT Reconstruction

On The Temporal Domain of Differential Equation Inspired Graph Neural Networks

no code implementations20 Jan 2024 Moshe Eliasof, Eldad Haber, Eran Treister, Carola-Bibiane Schönlieb

Graph Neural Networks (GNNs) have demonstrated remarkable success in modeling complex relationships in graph-structured data.

Second-order methods

The curious case of the test set AUROC

1 code implementation19 Dec 2023 Michael Roberts, Alon Hazan, Sören Dittmer, James H. F. Rudd, Carola-Bibiane Schönlieb

Whilst the size and complexity of ML models have rapidly and significantly increased over the past decade, the methods for assessing their performance have not kept pace.

Specificity

TrafficMOT: A Challenging Dataset for Multi-Object Tracking in Complex Traffic Scenarios

no code implementations30 Nov 2023 Lihao Liu, Yanqi Cheng, Zhongying Deng, Shujun Wang, Dongdong Chen, Xiaowei Hu, Pietro Liò, Carola-Bibiane Schönlieb, Angelica Aviles-Rivero

Multi-object tracking in traffic videos is a crucial research area, offering immense potential for enhancing traffic monitoring accuracy and promoting road safety measures through the utilisation of advanced machine learning algorithms.

Multi-Object Tracking Object

Closing the ODE-SDE gap in score-based diffusion models through the Fokker-Planck equation

no code implementations27 Nov 2023 Teo Deveney, Jan Stanczuk, Lisa Maria Kreusser, Chris Budd, Carola-Bibiane Schönlieb

In this paper we rigorously describe the range of dynamics and approximations that arise when training score-based diffusion models, including the true SDE dynamics, the neural approximations, the various approximate particle dynamics that result, as well as their associated Fokker--Planck equations and the neural network approximations of these Fokker--Planck equations.

Single-Shot Plug-and-Play Methods for Inverse Problems

no code implementations22 Nov 2023 Yanqi Cheng, Lipei Zhang, Zhenda Shen, Shujun Wang, Lequan Yu, Raymond H. Chan, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

In this work, we introduce Single-Shot PnP methods (SS-PnP), shifting the focus to solving inverse problems with minimal data.

TRIDENT: The Nonlinear Trilogy for Implicit Neural Representations

no code implementations21 Nov 2023 Zhenda Shen, Yanqi Cheng, Raymond H. Chan, Pietro Liò, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

Implicit neural representations (INRs) have garnered significant interest recently for their ability to model complex, high-dimensional data without explicit parameterisation.

Traffic Video Object Detection using Motion Prior

no code implementations16 Nov 2023 Lihao Liu, Yanqi Cheng, Dongdong Chen, Jing He, Pietro Liò, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

In this work, we propose two innovative methods to exploit the motion prior and boost the performance of both fully-supervised and semi-supervised traffic video object detection.

Object object-detection +1

Resilient Graph Neural Networks: A Coupled Dynamical Systems Approach

no code implementations12 Nov 2023 Moshe Eliasof, Davide Murari, Ferdia Sherry, Carola-Bibiane Schönlieb

Graph Neural Networks (GNNs) have established themselves as a key component in addressing diverse graph-based tasks.

Provably Convergent Data-Driven Convex-Nonconvex Regularization

no code implementations9 Oct 2023 Zakhar Shumaylov, Jeremy Budd, Subhadip Mukherjee, Carola-Bibiane Schönlieb

An emerging new paradigm for solving inverse problems is via the use of deep learning to learn a regularizer from data.

Riemannian geometry for efficient analysis of protein dynamics data

1 code implementation15 Aug 2023 Willem Diepeveen, Carlos Esteve-Yagüe, Jan Lellmann, Ozan Öktem, Carola-Bibiane Schönlieb

First, it comes with a rich structure to account for a wide range of geometries that can be modelled after an energy landscape.

Convergent regularization in inverse problems and linear plug-and-play denoisers

no code implementations18 Jul 2023 Andreas Hauptmann, Subhadip Mukherjee, Carola-Bibiane Schönlieb, Ferdia Sherry

While a significant amount of research has gone into establishing the convergence of the PnP iteration for different regularity conditions on the denoisers, not much is known about the asymptotic properties of the converged solution as the noise level in the measurement tends to zero, i. e., whether PnP methods are provably convergent regularization schemes under reasonable assumptions on the denoiser.

Denoising Image Reconstruction

Inverse Evolution Layers: Physics-informed Regularizers for Deep Neural Networks

no code implementations14 Jul 2023 Chaoyu Liu, Zhonghua Qiao, Chao Li, Carola-Bibiane Schönlieb

Traditional image processing methods employing partial differential equations (PDEs) offer a multitude of meaningful regularizers, along with valuable theoretical foundations for a wide range of image-related tasks.

Semantic Segmentation

Designing Stable Neural Networks using Convex Analysis and ODEs

1 code implementation29 Jun 2023 Ferdia Sherry, Elena Celledoni, Matthias J. Ehrhardt, Davide Murari, Brynjulf Owren, Carola-Bibiane Schönlieb

Motivated by classical work on the numerical integration of ordinary differential equations we present a ResNet-styled neural network architecture that encodes non-expansive (1-Lipschitz) operators, as long as the spectral norms of the weights are appropriately constrained.

Deblurring image-classification +3

CDiffMR: Can We Replace the Gaussian Noise with K-Space Undersampling for Fast MRI?

1 code implementation25 Jun 2023 Jiahao Huang, Angelica Aviles-Rivero, Carola-Bibiane Schönlieb, Guang Yang

Different from conventional diffusion models, the degradation operation of our CDiffMR is based on \textit{k}-space undersampling instead of adding Gaussian noise, and the restoration network is trained to harness a de-aliaseing function.

MRI Reconstruction

Variational Diffusion Auto-encoder: Latent Space Extraction from Pre-trained Diffusion Models

no code implementations24 Apr 2023 Georgios Batzolis, Jan Stanczuk, Carola-Bibiane Schönlieb

This issue stems from the unrealistic assumption that approximates the conditional data distribution, $p(\textbf{x} | \textbf{z})$, as an isotropic Gaussian.

Decoder

DiffMIC: Dual-Guidance Diffusion Network for Medical Image Classification

1 code implementation19 Mar 2023 Yijun Yang, Huazhu Fu, Angelica I. Aviles-Rivero, Carola-Bibiane Schönlieb, Lei Zhu

However, while a substantial amount of diffusion-based research has focused on generative tasks, few studies have applied diffusion models to general medical image classification.

Diabetic Retinopathy Grading image-classification +4

HGIB: Prognosis for Alzheimer's Disease via Hypergraph Information Bottleneck

no code implementations18 Mar 2023 Shujun Wang, Angelica I Aviles-Rivero, Zoe Kourtzi, Carola-Bibiane Schönlieb

We demonstrate, through extensive experiments on ADNI, that our proposed HGIB framework outperforms existing state-of-the-art hypergraph neural networks for Alzheimer's disease prognosis.

Prognosis

Class-Guided Image-to-Image Diffusion: Cell Painting from Brightfield Images with Class Labels

1 code implementation15 Mar 2023 Jan Oscar Cross-Zamirski, Praveen Anand, Guy Williams, Elizabeth Mouchet, Yinhai Wang, Carola-Bibiane Schönlieb

Image-to-image reconstruction problems with free or inexpensive metadata in the form of class labels appear often in biological and medical image domains.

Denoising Drug Discovery +2

CoNIC Challenge: Pushing the Frontiers of Nuclear Detection, Segmentation, Classification and Counting

1 code implementation11 Mar 2023 Simon Graham, Quoc Dang Vu, Mostafa Jahanifar, Martin Weigert, Uwe Schmidt, Wenhua Zhang, Jun Zhang, Sen yang, Jinxi Xiang, Xiyue Wang, Josef Lorenz Rumberger, Elias Baumann, Peter Hirsch, Lihao Liu, Chenyang Hong, Angelica I. Aviles-Rivero, Ayushi Jain, Heeyoung Ahn, Yiyu Hong, Hussam Azzuni, Min Xu, Mohammad Yaqub, Marie-Claire Blache, Benoît Piégu, Bertrand Vernay, Tim Scherr, Moritz Böhland, Katharina Löffler, Jiachen Li, Weiqin Ying, Chixin Wang, Dagmar Kainmueller, Carola-Bibiane Schönlieb, Shuolin Liu, Dhairya Talsania, Yughender Meda, Prakash Mishra, Muhammad Ridzuan, Oliver Neumann, Marcel P. Schilling, Markus Reischl, Ralf Mikut, Banban Huang, Hsiang-Chin Chien, Ching-Ping Wang, Chia-Yen Lee, Hong-Kun Lin, Zaiyi Liu, Xipeng Pan, Chu Han, Jijun Cheng, Muhammad Dawood, Srijay Deshpande, Raja Muhammad Saad Bashir, Adam Shephard, Pedro Costa, João D. Nunes, Aurélio Campilho, Jaime S. Cardoso, Hrishikesh P S, Densen Puthussery, Devika R G, Jiji C V, Ye Zhang, Zijie Fang, Zhifan Lin, Yongbing Zhang, Chunhui Lin, Liukun Zhang, Lijian Mao, Min Wu, Vi Thi-Tuong Vo, Soo-Hyung Kim, Taebum Lee, Satoshi Kondo, Satoshi Kasai, Pranay Dumbhare, Vedant Phuse, Yash Dubey, Ankush Jamthikar, Trinh Thi Le Vuong, Jin Tae Kwak, Dorsa Ziaei, Hyun Jung, Tianyi Miao, David Snead, Shan E Ahmed Raza, Fayyaz Minhas, Nasir M. Rajpoot

Nuclear detection, segmentation and morphometric profiling are essential in helping us further understand the relationship between histology and patient outcome.

Nuclear Segmentation Segmentation +2

Provably Convergent Plug-and-Play Quasi-Newton Methods

1 code implementation9 Mar 2023 Hong Ye Tan, Subhadip Mukherjee, Junqi Tang, Carola-Bibiane Schönlieb

Plug-and-Play (PnP) methods are a class of efficient iterative methods that aim to combine data fidelity terms and deep denoisers using classical optimization algorithms, such as ISTA or ADMM, with applications in inverse problems and imaging.

Deblurring Image Deblurring +1

Can Physics-Informed Neural Networks beat the Finite Element Method?

1 code implementation8 Feb 2023 Tamara G. Grossmann, Urszula Julia Komorowska, Jonas Latz, Carola-Bibiane Schönlieb

In terms of solution time and accuracy, physics-informed neural networks have not been able to outperform the finite element method in our study.

Continuous U-Net: Faster, Greater and Noiseless

no code implementations1 Feb 2023 Chun-Wun Cheng, Christina Runkel, Lihao Liu, Raymond H Chan, Carola-Bibiane Schönlieb, Angelica I Aviles-Rivero

Despite the powerful performance reported by existing U-Net type networks, they suffer from several major limitations.

Benchmarking Decoder +4

Your diffusion model secretly knows the dimension of the data manifold

1 code implementation23 Dec 2022 Jan Stanczuk, Georgios Batzolis, Teo Deveney, Carola-Bibiane Schönlieb

A diffusion model approximates the score function i. e. the gradient of the log density of a noise-corrupted version of the target distribution for varying levels of corruption.

Navigating the challenges in creating complex data systems: a development philosophy

no code implementations21 Oct 2022 Sören Dittmer, Michael Roberts, Julian Gilbey, Ander Biguri, AIX-COVNET Collaboration, Jacobus Preller, James H. F. Rudd, John A. D. Aston, Carola-Bibiane Schönlieb

In this perspective, we argue that despite the democratization of powerful tools for data science and machine learning over the last decade, developing the code for a trustworthy and effective data science system (DSS) is getting harder.

Philosophy

Dynamical systems' based neural networks

1 code implementation5 Oct 2022 Elena Celledoni, Davide Murari, Brynjulf Owren, Carola-Bibiane Schönlieb, Ferdia Sherry

The structure of the neural network is then inferred from the properties of the ODE vector field.

Self-Supervised Learning of Phenotypic Representations from Cell Images with Weak Labels

1 code implementation16 Sep 2022 Jan Oscar Cross-Zamirski, Guy Williams, Elizabeth Mouchet, Carola-Bibiane Schönlieb, Riku Turkki, Yinhai Wang

We propose WS-DINO as a novel framework to use weak label information in learning phenotypic representations from high-content fluorescent images of cells.

Knowledge Distillation Self-Supervised Learning

Spectral decomposition of atomic structures in heterogeneous cryo-EM

1 code implementation12 Sep 2022 Carlos Esteve-Yagüe, Willem Diepeveen, Ozan Öktem, Carola-Bibiane Schönlieb

The dataset contains noisy tomographic projections of the electrostatic potential of the macromolecule, taken from different viewing directions, and in the heterogeneous case, each image corresponds to a different conformation of the macromolecule.

Imaging with Equivariant Deep Learning

no code implementations5 Sep 2022 Dongdong Chen, Mike Davies, Matthias J. Ehrhardt, Carola-Bibiane Schönlieb, Ferdia Sherry, Julián Tachella

From early image processing to modern computational imaging, successful models and algorithms have relied on a fundamental property of natural signals: symmetry.

compressed sensing Deep Learning +3

Joint reconstruction-segmentation on graphs

no code implementations11 Aug 2022 Jeremy Budd, Yves van Gennip, Jonas Latz, Simone Parisotto, Carola-Bibiane Schönlieb

Practical image segmentation tasks concern images which must be reconstructed from noisy, distorted, and/or incomplete observations.

Image Segmentation Segmentation +1

Stochastic Primal-Dual Three Operator Splitting Algorithm with Extension to Equivariant Regularization-by-Denoising

no code implementations2 Aug 2022 Junqi Tang, Matthias Ehrhardt, Carola-Bibiane Schönlieb

In this work we propose a stochastic primal-dual three-operator splitting algorithm (TOS-SPDHG) for solving a class of convex three-composite optimization problems.

Denoising

Non-Uniform Diffusion Models

no code implementations20 Jul 2022 Georgios Batzolis, Jan Stanczuk, Carola-Bibiane Schönlieb, Christian Etmann

We show that non-uniform diffusion leads to multi-scale diffusion models which have similar structure to this of multi-scale normalizing flows.

Denoising

Unsupervised Learning of the Total Variation Flow

1 code implementation9 Jun 2022 Tamara G. Grossmann, Sören Dittmer, Yury Korolev, Carola-Bibiane Schönlieb

Inspired by and extending the framework of physics-informed neural networks (PINNs), we propose the TVflowNET, an unsupervised neural network approach, to approximate the solution of the TV flow given an initial image and a time instance.

Texture Classification

Multi-Modal Hypergraph Diffusion Network with Dual Prior for Alzheimer Classification

no code implementations4 Apr 2022 Angelica I. Aviles-Rivero, Christina Runkel, Nicolas Papadakis, Zoe Kourtzi, Carola-Bibiane Schönlieb

We demonstrate, through our experiments, that our framework is able to outperform current techniques for Alzheimer's disease diagnosis.

Multi-modal Classification

Operator Sketching for Deep Unrolling Networks

no code implementations21 Mar 2022 Junqi Tang, Subhadip Mukherjee, Carola-Bibiane Schönlieb

In this work we propose a new paradigm for designing efficient deep unrolling networks using operator sketching.

Image Reconstruction Rolling Shutter Correction

Multi-modal learning for predicting the genotype of glioma

no code implementations21 Mar 2022 Yiran Wei, Xi Chen, Lei Zhu, Lipei Zhang, Carola-Bibiane Schönlieb, Stephen J. Price, Chao Li

In this study, we propose a multi-modal learning framework using three separate encoders to extract features of focal tumor image, tumor geometrics and global brain networks.

Clinical Knowledge Diffusion MRI +1

Unsupervised Clustering of Roman Potsherds via Variational Autoencoders

no code implementations14 Mar 2022 Simone Parisotto, Ninetta Leone, Carola-Bibiane Schönlieb, Alessandro Launaro

In this paper we propose an artificial intelligence imaging solution to support archaeologists in the classification task of Roman commonware potsherds.

Clustering

Mutual Contrastive Low-rank Learning to Disentangle Whole Slide Image Representations for Glioma Grading

no code implementations8 Mar 2022 Lipei Zhang, Yiran Wei, Ying Fu, Stephen Price, Carola-Bibiane Schönlieb, Chao Li

In this proposed scheme, we design a normalized modality contrastive loss (NMC-loss), which could promote to disentangle multi-modality complementary representation of FFPE and frozen sections from the same patient.

Contrastive Learning Diagnostic +2

Predicting conversion of mild cognitive impairment to Alzheimer's disease

no code implementations8 Mar 2022 Yiran Wei, Stephen J. Price, Carola-Bibiane Schönlieb, Chao Li

In this study, we develop a self-supervised contrastive learning approach to generate structural brain networks from routine anatomical MRI under the guidance of diffusion MRI.

Contrastive Learning Diffusion MRI +2

Collaborative learning of images and geometrics for predicting isocitrate dehydrogenase status of glioma

no code implementations14 Jan 2022 Yiran Wei, Chao Li, Xi Chen, Carola-Bibiane Schönlieb, Stephen J. Price

Further, the collaborative learning model achieves better performance than either the CNN or the GNN alone.

A Continuous-time Stochastic Gradient Descent Method for Continuous Data

no code implementations7 Dec 2021 Kexin Jin, Jonas Latz, ChenGuang Liu, Carola-Bibiane Schönlieb

Optimization problems with continuous data appear in, e. g., robust machine learning, functional data analysis, and variational inference.

Stochastic Optimization Variational Inference

Conditional Image Generation with Score-Based Diffusion Models

1 code implementation26 Nov 2021 Georgios Batzolis, Jan Stanczuk, Carola-Bibiane Schönlieb, Christian Etmann

Score-based diffusion models have emerged as one of the most promising frameworks for deep generative modelling.

Conditional Image Generation

Calibrating the Dice loss to handle neural network overconfidence for biomedical image segmentation

1 code implementation31 Oct 2021 Michael Yeung, Leonardo Rundo, Yang Nan, Evis Sala, Carola-Bibiane Schönlieb, Guang Yang

However, it is well known that the DSC loss is poorly calibrated, resulting in overconfident predictions that cannot be usefully interpreted in biomedical and clinical practice.

Image Segmentation Segmentation +1

Focal Attention Networks: optimising attention for biomedical image segmentation

no code implementations31 Oct 2021 Michael Yeung, Leonardo Rundo, Evis Sala, Carola-Bibiane Schönlieb, Guang Yang

In recent years, there has been increasing interest to incorporate attention into deep learning architectures for biomedical image segmentation.

Image Segmentation Semantic Segmentation

Learning convex regularizers satisfying the variational source condition for inverse problems

no code implementations NeurIPS Workshop Deep_Invers 2021 Subhadip Mukherjee, Carola-Bibiane Schönlieb, Martin Burger

Variational regularization has remained one of the most successful approaches for reconstruction in imaging inverse problems for several decades.

Stochastic Primal-Dual Deep Unrolling

no code implementations19 Oct 2021 Junqi Tang, Subhadip Mukherjee, Carola-Bibiane Schönlieb

We develop a stochastic (ordered-subsets) variant of the classical learned primal-dual (LPD), which is a state-of-the-art unrolling network for tomographic image reconstruction.

Computational Efficiency Computed Tomography (CT) +2

StyleGAN-induced data-driven regularization for inverse problems

no code implementations7 Oct 2021 Arthur Conmy, Subhadip Mukherjee, Carola-Bibiane Schönlieb

Our proposed approach, which we refer to as learned Bayesian reconstruction with generative models (L-BRGM), entails joint optimization over the style-code and the input latent code, and enhances the expressive power of a pre-trained StyleGAN2 generator by allowing the style-codes to be different for different generator layers.

Image Inpainting Image Reconstruction +1

LaplaceNet: A Hybrid Graph-Energy Neural Network for Deep Semi-Supervised Classification

1 code implementation8 Jun 2021 Philip Sellars, Angelica I. Aviles-Rivero, Carola-Bibiane Schönlieb

Semi-supervised learning has received a lot of recent attention as it alleviates the need for large amounts of labelled data which can often be expensive, requires expert knowledge and be time consuming to collect.

Semi-Supervised Image Classification

HERS Superpixels: Deep Affinity Learning for Hierarchical Entropy Rate Segmentation

1 code implementation7 Jun 2021 Hankui Peng, Angelica I. Aviles-Rivero, Carola-Bibiane Schönlieb

Using the learned affinities from the first stage, HERS builds a hierarchical tree structure that can produce any number of highly adaptive superpixels instantaneously.

Superpixels

Focus U-Net: A novel dual attention-gated CNN for polyp segmentation during colonoscopy

no code implementations16 May 2021 Michael Yeung, Evis Sala, Carola-Bibiane Schönlieb, Leonardo Rundo

When evaluated on a combination of five public polyp datasets, our model similarly achieves state-of-the-art results with a mean DSC of 0. 878 and mean IoU of 0. 809, a 14% and 15% improvement over the previous state-of-the-art results of 0. 768 and 0. 702, respectively.

Image Segmentation Semantic Segmentation

An end-to-end Optical Character Recognition approach for ultra-low-resolution printed text images

no code implementations10 May 2021 Julian D. Gilbey, Carola-Bibiane Schönlieb

Our experiments have shown that it is possible to perform OCR on 60 dpi scanned images of English text, which is a significantly lower resolution than the state-of-the-art, and we achieved a mean character level accuracy (CLA) of 99. 7% and word level accuracy (WLA) of 98. 9% across a set of about 1000 pages of 60 dpi text in a wide range of fonts.

Optical Character Recognition Optical Character Recognition (OCR) +1

Semi-supervised Superpixel-based Multi-Feature Graph Learning for Hyperspectral Image Data

no code implementations27 Apr 2021 Madeleine Kotzagiannidis, Carola-Bibiane Schönlieb

In this work, we present a novel framework for the classification of HSI data in light of a very limited amount of labelled data, inspired by multi-view graph learning and graph signal processing.

graph construction Graph Learning +1

Adversarially learned iterative reconstruction for imaging inverse problems

1 code implementation30 Mar 2021 Subhadip Mukherjee, Ozan Öktem, Carola-Bibiane Schönlieb

In numerous practical applications, especially in medical image reconstruction, it is often infeasible to obtain a large ensemble of ground-truth/measurement pairs for supervised learning.

Image Reconstruction

Wasserstein GANs Work Because They Fail (to Approximate the Wasserstein Distance)

no code implementations2 Mar 2021 Jan Stanczuk, Christian Etmann, Lisa Maria Kreusser, Carola-Bibiane Schönlieb

Wasserstein GANs are based on the idea of minimising the Wasserstein distance between a real and a generated distribution.

Equivariant neural networks for inverse problems

1 code implementation23 Feb 2021 Elena Celledoni, Matthias J. Ehrhardt, Christian Etmann, Brynjulf Owren, Carola-Bibiane Schönlieb, Ferdia Sherry

In this work, we demonstrate that group equivariant convolutional operations can naturally be incorporated into learned reconstruction methods for inverse problems that are motivated by the variational regularisation approach.

Inductive Bias

Unified Focal loss: Generalising Dice and cross entropy-based losses to handle class imbalanced medical image segmentation

5 code implementations8 Feb 2021 Michael Yeung, Evis Sala, Carola-Bibiane Schönlieb, Leonardo Rundo

We compare our loss function performance against six Dice or cross entropy-based loss functions, across 2D binary, 3D binary and 3D multiclass segmentation tasks, demonstrating that our proposed loss function is robust to class imbalance and consistently outperforms the other loss functions.

Image Segmentation Medical Image Analysis +3

Bayesian optimization assisted unsupervised learning for efficient intra-tumor partitioning in MRI and survival prediction for glioblastoma patients

no code implementations5 Dec 2020 YiFan Li, Chao Li, Stephen Price, Carola-Bibiane Schönlieb, Xi Chen

Although successful in tumor sub-region segmentation and survival prediction, radiomics based on machine learning algorithms, is challenged by its robustness, due to the vague intermediate process and track changes.

Bayesian Optimization BIG-bench Machine Learning +4

A Three-Stage Self-Training Framework for Semi-Supervised Semantic Segmentation

1 code implementation1 Dec 2020 Rihuan Ke, Angelica Aviles-Rivero, Saurabh Pandey, Saikumar Reddy, Carola-Bibiane Schönlieb

The key idea of our technique is the extraction of the pseudo-masks statistical information to decrease uncertainty in the predicted probability whilst enforcing segmentation consistency in a multi-task fashion.

Segmentation Semi-Supervised Semantic Segmentation

TFPnP: Tuning-free Plug-and-Play Proximal Algorithm with Applications to Inverse Imaging Problems

1 code implementation18 Nov 2020 Kaixuan Wei, Angelica Aviles-Rivero, Jingwei Liang, Ying Fu, Hua Huang, Carola-Bibiane Schönlieb

In this work, we present a class of tuning-free PnP proximal algorithms that can determine parameters such as denoising strength, termination time, and other optimization-specific parameters automatically.

compressed sensing Deep Reinforcement Learning +2

Contrastive Registration for Unsupervised Medical Image Segmentation

1 code implementation17 Nov 2020 Lihao Liu, Angelica I Aviles-Rivero, Carola-Bibiane Schönlieb

Secondly, we embed a contrastive learning mechanism into the registration architecture to enhance the discriminating capacity of the network in the feature-level.

Contrastive Learning Image Segmentation +3

Regularized Compression of MRI Data: Modular Optimization of Joint Reconstruction and Coding

no code implementations8 Oct 2020 Veronica Corona, Yehuda Dar, Guy Williams, Carola-Bibiane Schönlieb

In this work we propose a framework for joint optimization of the MRI reconstruction and lossy compression, producing compressed representations of medical images that achieve improved trade-offs between quality and bit-rate.

Data Compression Image Compression +2

GraphXCOVID: Explainable Deep Graph Diffusion Pseudo-Labelling for Identifying COVID-19 on Chest X-rays

no code implementations30 Sep 2020 Angelica I. Aviles-Rivero, Philip Sellars, Carola-Bibiane Schönlieb, Nicolas Papadakis

The creation of which is a heavily expensive and time consuming task, and especially imposes a great challenge for a novel disease.

Diagnostic

A Linear Transportation $\mathrm{L}^p$ Distance for Pattern Recognition

no code implementations23 Sep 2020 Oliver M. Crook, Mihai Cucuringu, Tim Hurst, Carola-Bibiane Schönlieb, Matthew Thorpe, Konstantinos C. Zygalakis

The transportation $\mathrm{L}^p$ distance, denoted $\mathrm{TL}^p$, has been proposed as a generalisation of Wasserstein $\mathrm{W}^p$ distances motivated by the property that it can be applied directly to colour or multi-channelled images, as well as multivariate time-series without normalisation or mass constraints.

Time Series Time Series Analysis

Unsupervised Image Restoration Using Partially Linear Denoisers

1 code implementation14 Aug 2020 Rihuan Ke, Carola-Bibiane Schönlieb

The ground truth images, however, are often unavailable or very expensive to acquire in real-world applications.

Deblurring Image Denoising +1

Learned convex regularizers for inverse problems

1 code implementation6 Aug 2020 Subhadip Mukherjee, Sören Dittmer, Zakhar Shumaylov, Sebastian Lunz, Ozan Öktem, Carola-Bibiane Schönlieb

We consider the variational reconstruction framework for inverse problems and propose to learn a data-adaptive input-convex neural network (ICNN) as the regularization functional.

Computed Tomography (CT) Deblurring

Ground Truth Free Denoising by Optimal Transport

1 code implementation3 Jul 2020 Sören Dittmer, Carola-Bibiane Schönlieb, Peter Maass

We present a learned unsupervised denoising method for arbitrary types of data, which we explore on images and one-dimensional signals.

Denoising Generative Adversarial Network

Deeply Learned Spectral Total Variation Decomposition

1 code implementation NeurIPS 2020 Tamara G. Grossmann, Yury Korolev, Guy Gilboa, Carola-Bibiane Schönlieb

To the best of our knowledge, this is the first approach towards learning a non-linear spectral decomposition of images.

Structure preserving deep learning

no code implementations5 Jun 2020 Elena Celledoni, Matthias J. Ehrhardt, Christian Etmann, Robert I McLachlan, Brynjulf Owren, Carola-Bibiane Schönlieb, Ferdia Sherry

Over the past few years, deep learning has risen to the foreground as a topic of massive interest, mainly as a result of successes obtained in solving large-scale image processing tasks.

Deep Learning

Unsupervised clustering of Roman pottery profiles from their SSAE representation

no code implementations4 Jun 2020 Simone Parisotto, Alessandro Launaro, Ninetta Leone, Carola-Bibiane Schönlieb

The partiality and the handcrafted variance of the shape fragments within this new database make their unsupervised clustering a very challenging problem: profile similarities are thus explored via the hierarchical clustering of non-linear features learned in the latent representation space of a stacked sparse autoencoder (SSAE) network, unveiling new profile matches.

Clustering

Multi-task deep learning for image segmentation using recursive approximation tasks

no code implementations26 May 2020 Rihuan Ke, Aurélie Bugeau, Nicolas Papadakis, Mark Kirkland, Peter Schuetz, Carola-Bibiane Schönlieb

The subproblems are handled by a framework that consists of 1) a segmentation task that learns from pixel-level ground truth segmentation masks of a small fraction of the images, 2) a recursive approximation task that conducts partial object regions learning and data-driven mask evolution starting from partial masks of each object instance, and 3) other problem oriented auxiliary tasks that are trained with sparse annotations and promote the learning of dedicated features.

Image Segmentation Multi-Task Learning +3

3D deformable registration of longitudinal abdominopelvic CT images using unsupervised deep learning

no code implementations15 May 2020 Maureen van Eijnatten, Leonardo Rundo, K. Joost Batenburg, Felix Lucka, Emma Beddowes, Carlos Caldas, Ferdia A. Gallagher, Evis Sala, Carola-Bibiane Schönlieb, Ramona Woitek

This study showed the feasibility of deep learning based deformable registration of longitudinal abdominopelvic CT images via a novel incremental training strategy based on simulated deformations.

Image Registration

On Learned Operator Correction in Inverse Problems

1 code implementation14 May 2020 Sebastian Lunz, Andreas Hauptmann, Tanja Tarvainen, Carola-Bibiane Schönlieb, Simon Arridge

We discuss the possibility to learn a data-driven explicit model correction for inverse problems and whether such a model correction can be used within a variational framework to obtain regularised reconstructions.

iUNets: Fully invertible U-Nets with Learnable Up- and Downsampling

2 code implementations11 May 2020 Christian Etmann, Rihuan Ke, Carola-Bibiane Schönlieb

U-Nets have been established as a standard architecture for image-to-image learning problems such as segmentation and inverse problems in imaging.

SPRING: A fast stochastic proximal alternating method for non-smooth non-convex optimization

no code implementations27 Feb 2020 Derek Driggs, Junqi Tang, Jingwei Liang, Mike Davies, Carola-Bibiane Schönlieb

We introduce SPRING, a novel stochastic proximal alternating linearized minimization algorithm for solving a class of non-smooth and non-convex optimization problems.

Image Deconvolution Stochastic Optimization Optimization and Control 90C26

Total Variation Regularisation with Spatially Variable Lipschitz Constraints

1 code implementation5 Dec 2019 Martin Burger, Yury Korolev, Simone Parisotto, Carola-Bibiane Schönlieb

We introduce a first order Total Variation type regulariser that decomposes a function into a part with a given Lipschitz constant (which is also allowed to vary spatially) and a jump part.

Numerical Analysis Numerical Analysis 65J20, 65J22, 68U10, 94A08

Dynamic Spectral Residual Superpixels

no code implementations10 Oct 2019 Jianchao Zhang, Angelica I. Aviles-Rivero, Daniel Heydecker, Xiaosheng Zhuang, Raymond Chan, Carola-Bibiane Schönlieb

We consider the problem of segmenting an image into superpixels in the context of $k$-means clustering, in which we wish to decompose an image into local, homogeneous regions corresponding to the underlying objects.

Clustering Superpixels

A multi-task U-net for segmentation with lazy labels

no code implementations25 Sep 2019 Rihuan Ke, Aurélie Bugeau, Nicolas Papadakis, Peter Schuetz, Carola-Bibiane Schönlieb

The need for labour intensive pixel-wise annotation is a major limitation of many fully supervised learning methods for image segmentation.

Image Segmentation Multi-Task Learning +2

Learning the Sampling Pattern for MRI

2 code implementations20 Jun 2019 Ferdia Sherry, Martin Benning, Juan Carlos De los Reyes, Martin J. Graves, Georg Maierhofer, Guy Williams, Carola-Bibiane Schönlieb, Matthias J. Ehrhardt

The discovery of the theory of compressed sensing brought the realisation that many inverse problems can be solved even when measurements are "incomplete".

compressed sensing SSIM

On the Connection Between Adversarial Robustness and Saliency Map Interpretability

1 code implementation10 May 2019 Christian Etmann, Sebastian Lunz, Peter Maass, Carola-Bibiane Schönlieb

Recent studies on the adversarial vulnerability of neural networks have shown that models trained to be more robust to adversarial attacks exhibit more interpretable saliency maps than their non-robust counterparts.

Adversarial Robustness

Deep learning as optimal control problems: models and numerical methods

no code implementations11 Apr 2019 Martin Benning, Elena Celledoni, Matthias J. Ehrhardt, Brynjulf Owren, Carola-Bibiane Schönlieb

We review the first order conditions for optimality, and the conditions ensuring optimality after discretisation.

Deep Learning

Superpixel Contracted Graph-Based Learning for Hyperspectral Image Classification

1 code implementation14 Mar 2019 Philip Sellars, Angelica Aviles-Rivero, Carola-Bibiane Schönlieb

A central problem in hyperspectral image classification is obtaining high classification accuracy when using a limited amount of labelled data.

Classification General Classification +3

A total variation based regularizer promoting piecewise-Lipschitz reconstructions

no code implementations12 Mar 2019 Martin Burger, Yury Korolev, Carola-Bibiane Schönlieb, Christiane Stollenwerk

We introduce a new regularizer in the total variation family that promotes reconstructions with a given Lipschitz constant (which can also vary spatially).

Higher-Order Total Directional Variation. Part I: Imaging Applications

1 code implementation12 Dec 2018 Simone Parisotto, Jan Lellmann, Simon Masnou, Carola-Bibiane Schönlieb

We introduce a new class of higher-order total directional variation regularizers.

Numerical Analysis 47A52, 49M30, 49N45, 65J22, 94A08

Anisotropic osmosis filtering for shadow removal in images

1 code implementation17 Sep 2018 Simone Parisotto, Luca Calatroni, Marco Caliari, Carola-Bibiane Schönlieb, Joachim Weickert

We present an anisotropic extension of the isotropic osmosis model that has been introduced by Weickert et al.~(Weickert, 2013) for visual computing applications, and we adapt it specifically to shadow removal applications.

Analysis of PDEs 68U10, 94A08, 49K20, 65M06,

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

Faster PET Reconstruction with Non-Smooth Priors by Randomization and Preconditioning

1 code implementation21 Aug 2018 Matthias J. Ehrhardt, Pawel Markiewicz, Carola-Bibiane Schönlieb

Uncompressed clinical data from modern positron emission tomography (PET) scanners are very large, exceeding 350 million data points (projection bins).

Adversarial Regularizers in Inverse Problems

2 code implementations NeurIPS 2018 Sebastian Lunz, Ozan Öktem, Carola-Bibiane Schönlieb

Inverse Problems in medical imaging and computer vision are traditionally solved using purely model-based methods.

Denoising

Mirror, Mirror, on the Wall, Who's Got the Clearest Image of Them All? - A Tailored Approach to Single Image Reflection Removal

no code implementations29 May 2018 Daniel Heydecker, Georg Maierhofer, Angelica I. Aviles-Rivero, Qingnan Fan, Dong-Dong Chen, Carola-Bibiane Schönlieb, Sabine Süsstrunk

Removing reflection artefacts from a single image is a problem of both theoretical and practical interest, which still presents challenges because of the massively ill-posed nature of the problem.

All Reflection Removal

Unveiling the invisible - mathematical methods for restoring and interpreting illuminated manuscripts

1 code implementation19 Mar 2018 Luca Calatroni, Marie d'Autume, Rob Hocking, Stella Panayotova, Simone Parisotto, Paola Ricciardi, Carola-Bibiane Schönlieb

The last fifty years have seen an impressive development of mathematical methods for the analysis and processing of digital images, mostly in the context of photography, biomedical imaging and various forms of engineering.

Image Restoration

Blind Image Fusion for Hyperspectral Imaging with the Directional Total Variation

2 code implementations4 Oct 2017 Leon Bungert, David A. Coomes, Matthias J. Ehrhardt, Jennifer Rasch, Rafael Reisenhofer, Carola-Bibiane Schönlieb

In this paper, we propose a method for increasing the spatial resolution of a hyperspectral image by fusing it with an image of higher spatial resolution that was obtained with a different imaging modality.

Blind Super-Resolution Super-Resolution

Stochastic Primal-Dual Hybrid Gradient Algorithm with Arbitrary Sampling and Imaging Applications

2 code implementations15 Jun 2017 Antonin Chambolle, Matthias J. Ehrhardt, Peter Richtárik, Carola-Bibiane Schönlieb

We propose a stochastic extension of the primal-dual hybrid gradient algorithm studied by Chambolle and Pock in 2011 to solve saddle point problems that are separable in the dual variable.

Nonlinear Spectral Image Fusion

no code implementations23 Mar 2017 Martin Benning, Michael Möller, Raz Z. Nossek, Martin Burger, Daniel Cremers, Guy Gilboa, Carola-Bibiane Schönlieb

In this paper we demonstrate that the framework of nonlinear spectral decompositions based on total variation (TV) regularization is very well suited for image fusion as well as more general image manipulation tasks.

Image Manipulation

A Variational Model for Joint Motion Estimation and Image Reconstruction

no code implementations12 Jul 2016 Martin Burger, Hendrik Dirks, Carola-Bibiane Schönlieb

The aim of this paper is to derive and analyze a variational model for the joint estimation of motion and reconstruction of image sequences, which is based on a time-continuous Eulerian motion model.

Image Reconstruction Motion Estimation

The structure of optimal parameters for image restoration problems

no code implementations8 May 2015 Juan Carlos De Los Reyes, Carola-Bibiane Schönlieb, Tuomo Valkonen

The analysis is done on the original -- in image restoration typically non-smooth variational problem -- as well as on a smoothed approximation set in Hilbert space which is the one considered in numerical computations.

Image Restoration

Bilevel approaches for learning of variational imaging models

1 code implementation8 May 2015 Luca Calatroni, Cao Chung, Juan Carlos De Los Reyes, Carola-Bibiane Schönlieb, Tuomo Valkonen

We review some recent learning approaches in variational imaging, based on bilevel optimisation, and emphasize the importance of their treatment in function space.

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