About

I am currently a Computation & Theory Fellow at HHMI Janelia Research Campus.

With a multidisciplinary background in mathematics, computation, optics and biophysics, I am interested in various research questions at the intersection of machine learning, computer simulations, microscopy, and cell and tissue biology. Using computer simulations and machine learning tools, I aim to advance microscopy techniques and to gain insight into the workings of cells and tissues at the ultrastructure level. My interdisciplinary research has allowed me to gain broad experience collaborating with computer scientists, theorists and experimentalists.

If you are interested in my work, take a look at some of my projects below, or read more about my research at janelia.org.

Research Interests

I am interested in the following research topics:

  • Machine learning
  • Simulations
  • Neural representations
  • Graph neural networks
  • Adaptive optics
  • Super-resolution imaging

Projects

Here is a short overview of some of my projects:

Adaptive optics for fluorescence microscopy

Microscopy images are often distorted due to optical imperfections or inhomogeneities within sample. In collaboration with the Shroff lab, I work on enhancing wavefront sensing and image restoration using both classical and machine learning methods, with the aim of achieving optimal correction in real-time.
Analytical method: [Paper] [Code]
DeepPD: [Preprint]

Leveraging complementary microscopy methods

Every microscopy technique involves trade-offs between spatial and temporal resolution, contrast, and sample preservation. I develop and apply machine learning tools for image translation and integration, leveraging the strengths of multiple methods to achieve a more comprehensive understanding of cellular ultrastructure.

Simulation of point-spread functions

Fluorescent molecules exhibit anisotropic emission that leads to intricate point-spread functions. We developed a simulation tool to accurately model these PSFs, which is crucial to achieve accurate and unbiased results in single-molecule localization techniques.
[Paper] [Code]

Graph neural networks for understanding biological systems

Together with the Saalfeld lab at Janelia, we explore graph neural networks to learn the interaction rules between complex dynamical systems observed in biology.
[Preprint]

Publications

For a full list of publications please see my ORCID or Google Scholar page.

CVML Blog

We started a blog for our Janelia CVML interest group that highlights papers discussed by Janelia’s computer vision and machine learning researchers. I contributed to the following blog entries: