Welcome to Computational Imaging Systems Lab at Boston University!


loading slideshow...

  • Event-based Lightfield Microscopy, Light: Science & Applications (2024)

  • Spatially-varying FourierNet, Optica (2024)

  • DeepVID: Self-supervised denoising for voltage imaging, Nature Methods (2023)

  • Computational Miniature Mesoscope V2, Optica (2022)

  • Adaptive Learning by Dynamic Synthesis Network, Light: Science & Applications (2022)

  • Deep-learning augmented label-free multiplexed cytometry, Science Advances (2021)

  • Computational Miniature Mesoscope, Science Advances (2020)

  • Deep learning phase recovery with uncertainty quantification, Optica (2019)

  • High-speed intensity diffraction tomography, Advanced Photonics (2019)

  • Deep speckle correlation, Optica (2018)

We develop next-generation imaging systems that synergistically combine optics and computations. These computational imaging systems can overcome physical limitations and achieve novel capabilities that one could not with traditional imaging methods. The core of our work is the joint design of the physical layer (optical components, sensors, and devices) and the computational layer (signal processing, and machine learning algorithms).

We strive to develop innovative and impactful imaging techniques, with a broad spectrum of applications, in particular for biomedical microscopy, neuroscience, and semi-conductor metrology applications.

Please see our research page for more information.

We are grateful for the funding support from

nsf        nih