20 results on '"Scrivanti, Gabriele"'
Search Results
2. Forward-Backward algorithms for weakly convex problems
- Author
-
Bednarczuk, Ewa, Bruccola, Giovanni, Scrivanti, Gabriele, and Tran, The Hung
- Subjects
Mathematics - Optimization and Control ,90C30 90C26 90C51 65K10 52A01 - Abstract
We investigate the convergence properties of exact and inexact forward-backward algorithms to minimise the sum of two weakly convex functions defined on a Hilbert space, where one has a Lipschitz-continuous gradient. We show that the exact forward-backward algorithm converges strongly to a global solution, provided that the objective function satisfies a sharpness condition. For the inexact forward-backward algorithm, the same condition ensures that the distance from the iterates to the solution set approaches a positive threshold depending on the accuracy level of the proximal computations. As an application of the considered setting, we provide numerical experiments related to discrete tomography.
- Published
- 2023
3. Calculus rules for proximal {\epsilon}-subdifferentials and inexact proximity operators for weakly convex functions
- Author
-
Bednarczuk, Ewa, Bruccola, Giovanni, Scrivanti, Gabriele, and Tran, The Hung
- Subjects
Mathematics - Optimization and Control ,Mathematics - Numerical Analysis - Abstract
We investigate inexact proximity operators for weakly convex functions. To this aim, we derive sum rules for proximal {\epsilon}-subdifferentials, by incorporating the moduli of weak convexity of the functions into the respective formulas. This allows us to investigate inexact proximity operators for weakly convex functions in terms of proximal {\epsilon}-subdifferentials.
- Published
- 2022
4. A CNC approach for Directional Total Variation
- Author
-
Scrivanti, Gabriele, Chouzenoux, Emilie, and Pesquet, Jean-Christophe
- Subjects
Mathematics - Numerical Analysis - Abstract
The core of many approaches for the resolution of variational inverse problems arising in signal and image processing consists of promoting the sought solution to have a sparse representation in a well-suited space. A crucial task in this context is the choice of a good sparsity prior that can ensure a good trade-off between the quality of the solution and the resulting computational cost. The recently introduced Convex-Non-Convex (CNC) strategy appears as a great compromise, as it combines the high qualitative performance of non-convex sparsity-promoting functions with the convenience of dealing with convex optimization problems. This work proposes a new variational formulation to implement CNC approach in the context of image denoising. By suitably exploiting duality properties, our formulation allows to encompass sophisticated directional total variation (DTV) priors. We additionally propose an efficient optimisation strategy for the resulting convex minimisation problem. We illustrate on numerical examples the good performance of the resulting CNC-DTV method, when compared to the standard convex total variation denoiser., Comment: Accepted for EUSIPCO 2022 - 30th European Signal Processing Conference, Aug 2022, Belgrade, Serbia
- Published
- 2022
5. A Variational Approach for Joint Image Recovery and Feature Extraction Based on Spatially-Varying Generalised Gaussian Models
- Author
-
Chouzenoux, Emilie, Corbineau, Marie-Caroline, Pesquet, Jean-Christophe, and Scrivanti, Gabriele
- Subjects
Computer Science - Computer Vision and Pattern Recognition ,Mathematics - Numerical Analysis - Abstract
The joint problem of reconstruction / feature extraction is a challenging task in image processing. It consists in performing, in a joint manner, the restoration of an image and the extraction of its features. In this work, we firstly propose a novel nonsmooth and non-convex variational formulation of the problem. For this purpose, we introduce a versatile generalised Gaussian prior whose parameters, including its exponent, are space-variant. Secondly, we design an alternating proximal-based optimisation algorithm that efficiently exploits the structure of the proposed non-convex objective function. We also analyse the convergence of this algorithm. As shown in numerical experiments conducted on joint deblurring/segmentation tasks, the proposed method provides high-quality results.
- Published
- 2022
6. Non-convex Super-resolution of OCT images via sparse representation
- Author
-
Scrivanti, Gabriele, Calatroni, Luca, Morigi, Serena, Nicholson, Lindsay, and Achim, Alin
- Subjects
Electrical Engineering and Systems Science - Image and Video Processing ,Computer Science - Computer Vision and Pattern Recognition ,Mathematics - Numerical Analysis - Abstract
We propose a non-convex variational model for the super-resolution of Optical Coherence Tomography (OCT) images of the murine eye, by enforcing sparsity with respect to suitable dictionaries learnt from high-resolution OCT data. The statistical characteristics of OCT images motivate the use of {\alpha}-stable distributions for learning dictionaries, by considering the non-Gaussian case, {\alpha}=1. The sparsity-promoting cost function relies on a non-convex penalty - Cauchy-based or Minimax Concave Penalty (MCP) - which makes the problem particularly challenging. We propose an efficient algorithm for minimizing the function based on the forward-backward splitting strategy which guarantees at each iteration the existence and uniqueness of the proximal point. Comparisons with standard convex L1-based reconstructions show the better performance of non-convex models, especially in view of further OCT image analysis, Comment: 4 pages, 2 figures, 1 table, 1 algorithm, submitted to ISBI2021
- Published
- 2020
7. Space-variant image reconstruction via Cauchy regularisation: Application to Optical Coherence Tomography
- Author
-
Achim, Alin, Calatroni, Luca, Morigi, Serena, and Scrivanti, Gabriele
- Published
- 2023
- Full Text
- View/download PDF
8. Rapid and Noise‐Resilient Mapping of Photogenerated Carrier Lifetime in Halide Perovskite Thin Films.
- Author
-
Vidon, Guillaume, Scrivanti, Gabriele, Soret, Etienne, Harada, Nao, Chouzenoux, Emilie, Pesquet, Jean‐Christophe, Guillemoles, Jean‐François, and Cacovich, Stefania
- Subjects
- *
PEROVSKITE , *SOLAR energy , *THIN films , *PHOTOLUMINESCENCE , *OPTOELECTRONICS - Abstract
Halide perovskite materials offer significant promise for solar energy and optoelectronics yet understanding and enhancing their efficiency and stability require addressing lateral inhomogeneity challenges. While photoluminescence imaging techniques are employed for the measurement of their opto‐electronic and transport properties, going further in terms of precision requires longer acquisition times. Prolonged exposure of perovskites to light, given their high reactivity, can substantially alter these layers, rendering the acquired data less meaningful for analysis. In this paper, a method to extract high‐quality lifetime images from rapidly acquired, noisy time‐resolved photoluminescence images is proposed. This method leverages concepts of the field of constrained reconstruction and includes the Huber loss function and a specific form of total variation regularization. Through both simulations and experiments, it is demonstrated that the approach outperforms conventional pointwise methods. Optimal acceleration and optimization parameters tailored for decay time imaging of perovskite materials, offering new perspectives for accelerated experiments crucial in degradation process characterization are identified. Importantly, this methodology holds the potential for broader applications: it can be extended to explore additional beam‐sensitive materials, and other imaging characterization techniques and employed with more complex physical models to treat time‐resolved decays. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
9. Calculus rules for proximal ε-subdifferentials and inexact proximity operators for weakly convex functions
- Author
-
Bednarczuk, Ewa, primary, Bruccola, Giovanni, additional, Scrivanti, Gabriele, additional, and Tran, The Hung, additional
- Published
- 2023
- Full Text
- View/download PDF
10. Calculus rules for proximal ε-subdifferentials and inexact proximity operators for weakly convex functions
- Author
-
Bednarczuk, Ewa, Bruccola, Giovanni, Scrivanti, Gabriele, Tran, The Hung, Systems Research Institute [Warsaw] (IBS PAN), Polska Akademia Nauk = Polish Academy of Sciences (PAN), Warsaw University of Technology [Warsaw], OPtimisation Imagerie et Santé (OPIS), Inria Saclay - Ile de France, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre de vision numérique (CVN), Institut National de Recherche en Informatique et en Automatique (Inria)-CentraleSupélec-Université Paris-Saclay-CentraleSupélec-Université Paris-Saclay, Centre de vision numérique (CVN), Institut National de Recherche en Informatique et en Automatique (Inria)-CentraleSupélec-Université Paris-Saclay, European Project: 861137,TraDE-OPT(2020), Scrivanti, Gabriele, and Training Data-driven Experts in OPTimization - TraDE-OPT - 2020-06-01 - 2024-05-31 - 861137 - VALID
- Subjects
weakly convex functions ,Optimization and Control (math.OC) ,FOS: Mathematics ,[MATH.MATH-OC] Mathematics [math]/Optimization and Control [math.OC] ,inexactness ,Numerical Analysis (math.NA) ,proximal operator ,[MATH.MATH-OC]Mathematics [math]/Optimization and Control [math.OC] ,criticality ,sum rule for proximal ε-subdifferentials ,inexact proximal operator - Abstract
We investigate inexact proximity operators for weakly convex functions. To this aim, we derive sum rules for proximal ε-subdifferentials, by incorporating the moduli of weak convexity of the functions into the respective formulas. This allows us to investigate inexact proximity operators for weakly convex functions in terms of proximal ε-subdifferentials.
- Published
- 2022
11. Convergence analysis of an inexact Forward-Backward algorithm for problems involving weakly convex functions
- Author
-
Bednarczuk, Ewa, Bruccola, Giovanni, Scrivanti, Gabriele, and Tran, The Hung
- Subjects
Optimization and Control (math.OC) ,FOS: Mathematics ,90C30 90C26 65K10 52A01 ,Mathematics - Optimization and Control - Abstract
We investigate the convergence properties of exact and inexact forward-backward (FB) algorithms for the minimisation of the two functions $f+g$ defined on a Hilbert space, where $f$ is weakly convex and $g$ is convex and has a Lipschitz-continuous gradient. The main condition ensuring convergence is the sharpness of the objective $f+g$ around the solution set $S$. We show that the exact (FB) algorithm converges to a global solution provided that at a certain iteration the sequence is sufficiently close to the solution set. The inexact (FB) iterations converge to a ball around $S$ with radius depending upon the inexactness parameter $\varepsilon$. As an application of the analysed algorithm, we consider a feasibility problem involving a sphere and a closed convex set.
- Published
- 2023
12. A variational approach for joint image recovery-segmentation based on spatially varying generalised Gaussian models
- Author
-
Chouzenoux, Emilie, Corbineau, Marie-Caroline, Pesquet, Jean-Christophe, Scrivanti, Gabriele, Scrivanti, Gabriele, OPtimisation Imagerie et Santé (OPIS), Inria Saclay - Ile de France, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre de vision numérique (CVN), Institut National de Recherche en Informatique et en Automatique (Inria)-CentraleSupélec-Université Paris-Saclay-CentraleSupélec-Université Paris-Saclay, Centre de vision numérique (CVN), Institut National de Recherche en Informatique et en Automatique (Inria)-CentraleSupélec-Université Paris-Saclay, and This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 861137
- Subjects
Image recovery ,Space-variant regularisation ,Image segmentation ,Alternating minimization ,Texture decomposition ,Variable metric ,Proximal algorithm ,Block coordinate descent ,[MATH] Mathematics [math] ,[MATH]Mathematics [math] ,Kurdyka-Łojasiewicz property ,Ultrasound imaging AMS subject classifications - Abstract
The joint problem of reconstruction/feature extraction is a challenging task in image processing. It consists in performing, in a joint manner, the restoration of an image and the extraction of its features. In this work, we firstly propose a novel nonsmooth and nonconvex variational formulation of the problem. For this purpose, we introduce a versatile generalised Gaussian prior whose parameters, including its exponents, are space-variant. Secondly, we design an alternating proximal-based optimisation algorithm that efficiently exploits the structure of the proposed nonconvex objective function. We also analyze the convergence of this algorithm. As shown in numerical experiments conducted on joint segmentation/deblurring tasks, the proposed method provides high-quality results.
- Published
- 2022
13. A CNC approach for Directional Total Variation
- Author
-
Scrivanti, Gabriele, primary, Chouzenoux, Emilie, additional, and Pesquet, Jean-Christophe, additional
- Published
- 2022
- Full Text
- View/download PDF
14. Space-Variant Image Reconstruction Via Cauchy Regularisation: Application to Optical Coherence Tomography
- Author
-
Achim, Alin, primary, Calatroni, Luca, additional, Morigi, Serena, additional, and Scrivanti, Gabriele, additional
- Published
- 2022
- Full Text
- View/download PDF
15. Non-Convex Super-Resolution Of Oct Images Via Sparse Representation
- Author
-
Scrivanti, Gabriele, primary, Calatroni, Luca, additional, Morigi, Serena, additional, Nicholson, Lindsay, additional, and Achim, Alin, additional
- Published
- 2021
- Full Text
- View/download PDF
16. Nonsmooth Nonconvex Variational Reconstruction for Electrical Impedance Tomography
- Author
-
Scrivanti, Gabriele Luca Giovanni, thesis supervisor: Morigi, Serena, Scrivanti, Gabriele Luca Giovanni, and thesis supervisor: Morigi, Serena
- Abstract
Electrical Impedance Tomography is an imaging technique that aims to reconstruct the inner conductivity distribution of a medium starting from a set of measured voltages registered by a series of electrodes that are positioned on the surface of the medium. Such technique was used for the first time in geological studies in 1930 and then applied to industrial procedures. The first clinical use of EIT dates back to 1987. In 2018 EIT was validated in tissue engineering as a noninvasive and label-free imaging and monitoring tool for cell distribution (cell growth, differentiation and tissue formation) in 3D scaffolds. EIT problem can be split into a Forward and an Inverse problem. The aim of Forward EIT is to define the set of measured voltages starting from a known conductivity distribution. If the forward problem is characterized by a nonlinear mapping, called Forward Operator, from the conductivity distribution to the measured voltages, inverse EIT consists of inverting the Forward Operator. This leads to an ill-posed problem which requires regularization, either in the model or in the numerical method that is applied to define the solution. The inverse problem is modelled as a Nonlinear Least Squares problem, where one seeks to minimize the mismatch beetween the measured voltages and the ones generated by the reconstructed conductivity. Reconstruction techniques require the introduction of a regularization term which forces the reconstructed data to stick to certain prior information. In this dissertation, some state-of-the-art regularization methods are analyzed and compared via EIDORS, a specific software for EIT problems. The aim is to reconstruct the variation in conductivity within a 2D section of a 3D scaffold. Furthermore a variational formulation on a 2D mesh for a space-variant regularization is proposed, based on a combination of high order and nonconvex operators, which respectively seek to recover piecewise inhomogeneous and piecewise linear regions.
17. Due algoritmi per la fattorizzazione matriciale non negativa con applicazione al clustering
- Author
-
Scrivanti, Gabriele Luca Giovanni, thesis supervisor: Simoncini, Valeria, Scrivanti, Gabriele Luca Giovanni, and thesis supervisor: Simoncini, Valeria
- Abstract
In questa tesi vengono descritti il problema della fattorizzazione non negativa ortogonale (ONMF) con applicazione al clustering e due algoritmi elaborati da F. Pompili, N. Gillis, P.A. Absil e F. Gilneur per l'approssimazione numerica della coppia di matrici soluzione di tale problema: il primo algoritmo legato a una variante delle k-medie sferiche e il secondo basato sul metodo della Lagrangiana aumentata. Particolare attenzione viene prestata alla base teorica su cui si fonda il primo algoritmo, cioè l'equivalenza tra il problema delle k-medie sferiche pesate e il problema ONMF descritta dal Teorema 2.6. Per ciascun algoritmo vengono analizzati punti di forza e di debolezza e suggerita la tipologia di data set per cui risultano più indicati al fine di determinare un'opportuna divisione in cluster. Il primo capitolo, di carattere introduttivo, descrive i concetti di clustering e di fattorizzazione non negativa, proponendo una formulazione matematica utile ai fini della trattazione. Il secondo capitolo è dedicato all'algoritmo EMONMF, di cui è proposta la descrizione teorica e l'applicazione al problema di text clustering con oggetto una matrice termine-documento di articoli medici. Il terzo capitolo è dedicato all'algoritmo ONPMF di cui sono descritti i metodi di ottimizzazione su cui è costruito, cioè il metodo del gradiente proiettato e della Lagrangiana aumentata, e gli esperimenti numerici sono applicati all'Iris data set contenuto nel file matlab fisheriris. Infine, nel quarto e ultimo capitolo vengono proposti due confronti numerici degli algoritmi, che vengono analizzati in termini di iterazioni, tempi di elaborazione, stabilità e precisione della fattorizzazione e del clustering. Il primo confronto è applicato all'hyperspectral unmixing con oggetto il data set Hubble e il secondo è applicato al pattern recognition con oggetto U.S. Postal Service database. I codici Matlab sono disponibili all'indirizzo https://github.com/filippo-p/onmf.
18. Nonsmooth Nonconvex Variational Reconstruction for Electrical Impedance Tomography
- Author
-
Scrivanti, Gabriele Luca Giovanni, thesis supervisor: Morigi, Serena, Scrivanti, Gabriele Luca Giovanni, and thesis supervisor: Morigi, Serena
- Abstract
Electrical Impedance Tomography is an imaging technique that aims to reconstruct the inner conductivity distribution of a medium starting from a set of measured voltages registered by a series of electrodes that are positioned on the surface of the medium. Such technique was used for the first time in geological studies in 1930 and then applied to industrial procedures. The first clinical use of EIT dates back to 1987. In 2018 EIT was validated in tissue engineering as a noninvasive and label-free imaging and monitoring tool for cell distribution (cell growth, differentiation and tissue formation) in 3D scaffolds. EIT problem can be split into a Forward and an Inverse problem. The aim of Forward EIT is to define the set of measured voltages starting from a known conductivity distribution. If the forward problem is characterized by a nonlinear mapping, called Forward Operator, from the conductivity distribution to the measured voltages, inverse EIT consists of inverting the Forward Operator. This leads to an ill-posed problem which requires regularization, either in the model or in the numerical method that is applied to define the solution. The inverse problem is modelled as a Nonlinear Least Squares problem, where one seeks to minimize the mismatch beetween the measured voltages and the ones generated by the reconstructed conductivity. Reconstruction techniques require the introduction of a regularization term which forces the reconstructed data to stick to certain prior information. In this dissertation, some state-of-the-art regularization methods are analyzed and compared via EIDORS, a specific software for EIT problems. The aim is to reconstruct the variation in conductivity within a 2D section of a 3D scaffold. Furthermore a variational formulation on a 2D mesh for a space-variant regularization is proposed, based on a combination of high order and nonconvex operators, which respectively seek to recover piecewise inhomogeneous and piecewise linear regions.
19. Due algoritmi per la fattorizzazione matriciale non negativa con applicazione al clustering
- Author
-
Scrivanti, Gabriele Luca Giovanni, thesis supervisor: Simoncini, Valeria, Scrivanti, Gabriele Luca Giovanni, and thesis supervisor: Simoncini, Valeria
- Abstract
In questa tesi vengono descritti il problema della fattorizzazione non negativa ortogonale (ONMF) con applicazione al clustering e due algoritmi elaborati da F. Pompili, N. Gillis, P.A. Absil e F. Gilneur per l'approssimazione numerica della coppia di matrici soluzione di tale problema: il primo algoritmo legato a una variante delle k-medie sferiche e il secondo basato sul metodo della Lagrangiana aumentata. Particolare attenzione viene prestata alla base teorica su cui si fonda il primo algoritmo, cioè l'equivalenza tra il problema delle k-medie sferiche pesate e il problema ONMF descritta dal Teorema 2.6. Per ciascun algoritmo vengono analizzati punti di forza e di debolezza e suggerita la tipologia di data set per cui risultano più indicati al fine di determinare un'opportuna divisione in cluster. Il primo capitolo, di carattere introduttivo, descrive i concetti di clustering e di fattorizzazione non negativa, proponendo una formulazione matematica utile ai fini della trattazione. Il secondo capitolo è dedicato all'algoritmo EMONMF, di cui è proposta la descrizione teorica e l'applicazione al problema di text clustering con oggetto una matrice termine-documento di articoli medici. Il terzo capitolo è dedicato all'algoritmo ONPMF di cui sono descritti i metodi di ottimizzazione su cui è costruito, cioè il metodo del gradiente proiettato e della Lagrangiana aumentata, e gli esperimenti numerici sono applicati all'Iris data set contenuto nel file matlab fisheriris. Infine, nel quarto e ultimo capitolo vengono proposti due confronti numerici degli algoritmi, che vengono analizzati in termini di iterazioni, tempi di elaborazione, stabilità e precisione della fattorizzazione e del clustering. Il primo confronto è applicato all'hyperspectral unmixing con oggetto il data set Hubble e il secondo è applicato al pattern recognition con oggetto U.S. Postal Service database. I codici Matlab sono disponibili all'indirizzo https://github.com/filippo-p/onmf.
20. Non-Convex Super-Resolution Of Oct Images Via Sparse Representation
- Author
-
Lindsay B. Nicholson, Luca Calatroni, Alin Achim, Gabriele Scrivanti, Serena Morigi, Dipartimento di Matematica [Bologna], Alma Mater Studiorum Università di Bologna [Bologna] (UNIBO), Morphologie et Images (MORPHEME), Inria Sophia Antipolis - Méditerranée (CRISAM), Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Institut de Biologie Valrose (IBV), Université Nice Sophia Antipolis (... - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Université Côte d'Azur (UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Nice Sophia Antipolis (... - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Université Côte d'Azur (UCA)-Centre National de la Recherche Scientifique (CNRS)-Signal, Images et Systèmes (Laboratoire I3S - SIS), Laboratoire d'Informatique, Signaux, et Systèmes de Sophia Antipolis (I3S), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Université Nice Sophia Antipolis (... - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Laboratoire d'Informatique, Signaux, et Systèmes de Sophia Antipolis (I3S), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Centre National de la Recherche Scientifique (CNRS), University of Bristol [Bristol], Université Nice Sophia Antipolis (1965 - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Université Nice Sophia Antipolis (1965 - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Signal, Images et Systèmes (Laboratoire I3S - SIS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Université Nice Sophia Antipolis (1965 - 2019) (UNS), Calatroni, Luca, Scrivanti Gabriele, Luca Calatroni, Serena Morigi, L. Nicholson, and A. Achim
- Subjects
FOS: Computer and information sciences ,Computer science ,Computer Vision and Pattern Recognition (cs.CV) ,Computer Science - Computer Vision and Pattern Recognition ,MathematicsofComputing_NUMERICALANALYSIS ,ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION ,02 engineering and technology ,Image (mathematics) ,Optical coherence tomography ,FOS: Electrical engineering, electronic engineering, information engineering ,FOS: Mathematics ,0202 electrical engineering, electronic engineering, information engineering ,medicine ,[MATH.MATH-AP]Mathematics [math]/Analysis of PDEs [math.AP] ,Mathematics - Numerical Analysis ,Uniqueness ,[MATH.MATH-AP] Mathematics [math]/Analysis of PDEs [math.AP] ,medicine.diagnostic_test ,Image and Video Processing (eess.IV) ,Regular polygon ,Optical Coherence Tomography, Super-Resolution, Non-Convex Regularisation, Sparse Representation ,[MATH.MATH-OC] Mathematics [math]/Optimization and Control [math.OC] ,020206 networking & telecommunications ,Numerical Analysis (math.NA) ,Sparse approximation ,Function (mathematics) ,[MATH.MATH-NA] Mathematics [math]/Numerical Analysis [math.NA] ,Electrical Engineering and Systems Science - Image and Video Processing ,Minimax ,[INFO.INFO-TI] Computer Science [cs]/Image Processing [eess.IV] ,[INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV] ,020201 artificial intelligence & image processing ,[MATH.MATH-OC]Mathematics [math]/Optimization and Control [math.OC] ,Algorithm ,[MATH.MATH-NA]Mathematics [math]/Numerical Analysis [math.NA] ,Coherence (physics) - Abstract
We propose a non-convex variational model for the super-resolution of Optical Coherence Tomography (OCT) images of the murine eye, by enforcing sparsity with respect to suitable dictionaries learnt from high-resolution OCT data. The statistical characteristics of OCT images motivate the use of {\alpha}-stable distributions for learning dictionaries, by considering the non-Gaussian case, {\alpha}=1. The sparsity-promoting cost function relies on a non-convex penalty - Cauchy-based or Minimax Concave Penalty (MCP) - which makes the problem particularly challenging. We propose an efficient algorithm for minimizing the function based on the forward-backward splitting strategy which guarantees at each iteration the existence and uniqueness of the proximal point. Comparisons with standard convex L1-based reconstructions show the better performance of non-convex models, especially in view of further OCT image analysis, Comment: 4 pages, 2 figures, 1 table, 1 algorithm, submitted to ISBI2021
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.