I am a researcher in computational imaging with background in inverse problems, signal processing, and machine learning. My research focuses on developing robust, adaptive methods for solving ill-posed inverse problems where data is limited, noisy, or imperfectly modeled.

I really enjoy diving into new topics and expanding my horizons in new applications.

My background/Expertise

• • Superresolution and sparsity-driven reconstruction in inverse problems

  • Convex and non-convex optimization algorithms

  • Vector field tomography and imaging with singularities

  • Multigrid and iterative methods for large-scale inverse systems

  • Data assimilation and forecasting in dynamical systems

  • Machine learning for uncertainty quantification and model correction

  • Bayesian statistcs in inverse problems

  • Finite element methods for PDEs

Current Research Direction: 

• Learning Techniques for Uncertainty Reduction
  Applying data-driven methods to compensate for modeling errors and reduce uncertainties in inverse problems, with a focus on electroencephalographic (EEG) and magnetoencephalographic (MEG) brain imaging.
• Learning-Based Prior Parameter Estimation
  Investigating machine learning strategies for tuning regularization and estimating hyperparameters in statistical inversion, particularly in Poisson inverse problems and transport-reaction equation.

Previous problems that I have been involved:

• Superresolution in Linear Inverse Problems
  Developing adaptive reconstruction techniques using sparsity-promoting constraints to enhance spatial resolution in deconvolution and imaging systems. Applications include fluorescence microscopy and spectral feature extraction in spectroscopy
• Data Assimilation and Forecasting
  Integrating sparse and dynamic data into predictive models using assimilation frameworks, with applications ranging from ionospheric imaging to fault-tolerant satellite communication diagnostics.
• Vector Field Reconstruction via Tomographic Methods
  Designing tomographic approaches for reconstructing vector fields with singularities from limited measurement configurations in biomedical imaging.

 Some applications

• Neuroimaging: EEG/MEG source localization using Bayesian and optimization-based inverse modeling; mapping focal cortical activity.

• Biomedical Imaging: Deconvolution and superresolution in fluorescence microscopy and spectroscopy.

• Remote Sensing & Geophysics: ionospheric  imaging to visualize and assess scintillation impact on GNSS systems.