Software.

A PyTorch-based open-source library for generating accurate 2D and 3D point spread functions of high numerical-aperture microscopes, supporting both scalar and vectorial models. It enables GPU-accelerated computation and seamless integration with modern deep-learning workflows, with full documentation and Jupyter notebook demos.

An interactive napari plugin built on top of the PSF Generator library, allowing users to simulate and visualize point spread functions directly within the napari ecosystem. It supports both CPU and GPU backends, making it accessible for interactive microscopy design and parameter exploration.

Code companion to the paper on perturbative Fourier ptychographic microscopy (pFPM), a fast and high-resolution quantitative phase imaging method using LED array illumination. It implements the perturbative phase retrieval algorithm and the tailored annular dark-field illumination patterns described in the Optics Express 2025 paper.

A scikit-learn-compatible Python library for Echo State Networks and Reservoir Computing, designed to support both simulation and optical hardware implementations. It provides flexible reservoir architectures with support for encoding, structured transforms, and chaotic time-series prediction benchmarks (e.g., Mackey-Glass, Kuramoto-Sivashinsky).

A Jupyter notebook implementation of the spectral method for multiplexed phase retrieval, where phase information is recovered from intensity measurements that are linear combinations of multiple incoherent sources. It accompanies the ICASSP 2019 paper and demonstrates applications to light focusing in complex scattering media.

Code for the structured random model for phase retrieval, which emulates dense i.i.d. random matrices using a cascade of fast Fourier transforms and random diagonal matrices. It achieves log-linear computational complexity with no loss in reconstruction performance, as described in the IEEE ICASSP 2025 paper.

Code companion to the Optics Letters 2024 paper on Fisher-information-optimal PSF engineering for MINFLUX localization microscopy. It computes the optimal excitation patterns (donut and half-moon beams) that maximize localization precision using the Cramér-Rao bound framework.

Implementation of the automatic calibration algorithm introduced in Optics Continuum 2022, which systematically identifies and corrects mechanical artifacts in 3D OPT reconstructions. It includes a comprehensive catalog of artifact types along with an iterative reconstruction method that recovers unknown system parameters for distortion-free volumetric imaging.

Code for the cone-beam forward model and correction algorithm for refractive-index mismatch artifacts in OPT, described in Optics Letters 2022. It models the diverging-lens effect at the sample boundary and provides a fast reconstruction method validated on fluorescent bead phantoms.