Kyriakos Skarsoulis
Research Subjects
Kyriakos Skarsoulis has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Konstantinos G. Makris
Department of Physics, University of Crete, Heraklion, Greece
2 shared publications - Christophe Moser
Laboratory of Applied Photonics Devices, School of Engineering, École Polytechnique Fédérale de Lausanne, CH-1015, Lausanne, Switzerland
1 shared publication Demetri Psaltis
Institute of Electrical and Micro Engineering
1 shared publication- Demetri Psaltis
Optics Laboratory (LO), École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland
1 shared publication
Latest Publications
- Open Access
- 1 March 2026
- Published by AIP Publishing in APL Photonics
- Vol. 11 (3)
Deep learning in PT-symmetric multimode waveguide sensors
We examine the dynamic response of a waveguide with a PT-symmetric complex potential to perturbations in its refractive index. The output transverse intensity profile is recorded as different index perturbations are imposed. The waveguide exhibits its highest sensitivity when it is operating near its exceptional point (EP). A similar behavior is observed when multiple waveguides are coupled and operated around the new EP. A neural network is deployed to decode the cross-sectional intensity images and classify the strengths and positions of the applied perturbations. The proposed waveguide’s behavior near the EP allows the neural network to characterize the perturbations with high accuracy despite noise augmentation, contrary to the Hermitian case. Beyond 1D profiles, this scheme can be readily extended to recover full 2D perturbation distributions. A 1D PT-symmetric lattice structure comprised of five coupled waveguides is able to train a deep neural network capable of reconstructing a 2D perturbation map within the lattice from noisy intensity data, contrary to the Hermitian system. The results show how PT symmetry can be utilized in waveguides to create efficient sensors and imaging devices.
- 1 March 2026
- Published by AIP Publishing in APL Photonics
- Vol. 11 (3)
Open AccessDeep learning in PT-symmetric multimode waveguide sensors
We examine the dynamic response of a waveguide with a PT-symmetric complex potential to perturbations in its refractive index. The output transverse intensity profile is recorded as different index perturbations are imposed. The waveguide exhibits its highest sensitivity when it is operating near its exceptional point (EP). A similar behavior is observed when multiple waveguides are coupled and operated around the new EP. A neural network is deployed to decode the cross-sectional intensity images and classify the strengths and positions of the applied perturbations. The proposed waveguide’s behavior near the EP allows the neural network to characterize the perturbations with high accuracy despite noise augmentation, contrary to the Hermitian case. Beyond 1D profiles, this scheme can be readily extended to recover full 2D perturbation distributions. A 1D PT-symmetric lattice structure comprised of five coupled waveguides is able to train a deep neural network capable of reconstructing a 2D perturbation map within the lattice from noisy intensity data, contrary to the Hermitian system. The results show how PT symmetry can be utilized in waveguides to create efficient sensors and imaging devices.
- 30 May 2024
- Published by Optica Publishing Group in Optica
- Vol. 11 (6) , 782-790
Ptychographic imaging with a fiber endoscope via wavelength scanning
Ptychography has become a popular computational imaging method for microscopy in recent years. In the present work we employ a wavelength scanning ptychography technique enhanced by neural networks for imaging with a fiber endoscope. Illumination of the object at various wavelengths is achieved using a single mode fiber, while a multicore fiber collects diffracted light from a distance. Using a U-Net multilayer convolutional neural network, the diffraction pattern is recovered at the far end of the multicore fiber from the recorded intensity pattern at the proximal end. With the recovered diffraction pattern in place, the phase object can be reconstructed using the ptychography algorithm. The quality of the object reconstruction improves with the number of wavelengths used. Comparison with an end-to-end neural network highlights the effectiveness and practicality of this two-step hybrid system. This alternative and simplified ptychographic endoscopy setup delivers noticeable improvements through neural networks and wavelength scanning.
- 30 May 2024
- Published by Optica Publishing Group in Optica
- Vol. 11 (6) , 782-790
Ptychographic imaging with a fiber endoscope via wavelength scanning
Ptychography has become a popular computational imaging method for microscopy in recent years. In the present work we employ a wavelength scanning ptychography technique enhanced by neural networks for imaging with a fiber endoscope. Illumination of the object at various wavelengths is achieved using a single mode fiber, while a multicore fiber collects diffracted light from a distance. Using a U-Net multilayer convolutional neural network, the diffraction pattern is recovered at the far end of the multicore fiber from the recorded intensity pattern at the proximal end. With the recovered diffraction pattern in place, the phase object can be reconstructed using the ptychography algorithm. The quality of the object reconstruction improves with the number of wavelengths used. Comparison with an end-to-end neural network highlights the effectiveness and practicality of this two-step hybrid system. This alternative and simplified ptychographic endoscopy setup delivers noticeable improvements through neural networks and wavelength scanning.