A5087878038- Digital Holography and Microscopy
- Advanced Fluorescence Microscopy Techniques
- Random lasers and scattering media
- Advanced Vision and Imaging
- Advanced Data Storage Technologies
- Optical Coherence Tomography Applications
- Photoacoustic and Ultrasonic Imaging
- Image Processing Techniques and Applications
- Cloud Data Security Solutions
- Computer Graphics and Visualization Techniques
- Advanced Neural Network Applications
- Advanced Image Processing Techniques
- Near-Field Optical Microscopy
- Cell Image Analysis Techniques
- High-pressure geophysics and materials
- Diamond and Carbon-based Materials Research
- Optical Imaging and Spectroscopy Techniques
- Atomic and Subatomic Physics Research
Microsoft Research (United Kingdom)
2023-2024
University of California, Berkeley
2019-2023
Miniature fluorescence microscopes are a standard tool in systems biology. However, widefield miniature capture only 2D information, and modifications that enable 3D capabilities increase the size weight have poor resolution outside narrow depth range. Here, we achieve capability by replacing tube lens of conventional Miniscope with an optimized multifocal phase mask at objective's aperture stop. Placing stop significantly reduces device, varying focal lengths enables uniform across wide The...
We present an on-chip, widefield fluorescence microscope, which consists of a diffuser placed few millimeters away from traditional image sensor. The replaces the optics resulting in compact and easy-to-assemble system with practical working distance over 1.5 mm. Furthermore, encodes volumetric information, enabling refocusability post-processing three-dimensional (3D) imaging sparse samples single acquisition. Reconstruction images raw data requires precise model system, so we introduce...
Significance We report on progress toward combining magnetic resonance imaging (MRI) and optical in diamond microparticles. Our approach relies the nitrogen vacancy (NV) center-driven nuclear hyperpolarizability particles, that renders them “bright” MRI while simultaneously fluorescing optically. Both modes allow suppression of background signals. elucidate how such “dual-mode” can perform better than either taken separately, leveraging relative strengths optics with respect to resolution...
Sustainable and cost-effective long-term storage remains an unsolved problem. The most widely used technologies today are magnetic (hard disk drives tape). They use media that degrades over time has a limited lifetime, which leads to inefficient, wasteful, costly solutions for long-lived data. This paper presents Silica: the first cloud system archival data underpinned by quartz glass, extremely resilient allows be left in situ indefinitely. hardware software of Silica have been co-designed...
Light field microscopy (LFM) uses a microlens array (MLA) near the sensor plane of microscope to achieve single-shot 3D imaging sample without any moving parts. Unfortunately, capability LFM comes with significant loss lateral resolution at focal plane. Placing MLA pupil microscope, instead image plane, can mitigate artifacts and provide an efficient forward model, expense field-of-view (FOV). Here, we demonstrate improved across large volume Fourier DiffuserScope, which diffuser in encode...
Structured illumination microscopy (SIM) reconstructs a super-resolved image from multiple raw images captured with different patterns; hence, acquisition speed is limited, making it unsuitable for dynamic scenes. We propose new method, Speckle Flow SIM, that uses static patterned moving samples and models the sample motion during data capture in order to reconstruct scene super-resolution. SIM relies on sequence of images. The spatio-temporal relationship modeled using neural space-time...
We propose a single-shot 3D fluorescence microscope that achieves large field-of-view and good resolution across wide axial range. This is implemented by inserting diffuser into the Fourier plane of objective solving an inverse problem. show experimental results freely-moving C. elegans nematode at 25 fps.
Many computational imaging inverse problems are challenged by noise, model mismatch, and other imperfections that decrease reconstruction quality. For data taken sequentially in time, instead of reconstructing each frame independently, space-time algorithms simultaneously reconstruct multiple frames, thereby taking advantage temporal redundancy through priors. This helps with denoising provides improved quality, but often requires significant memory resources. Designing effective flexible...
Sustainable and cost-effective long-term storage remains an unsolved problem. The most widely used technologies today are magnetic (hard disk drives tape). They use media that degrades over time has a limited lifetime, which leads to inefficient, wasteful, costly solutions for long-lived data. This paper presents Silica: the first cloud system archival data underpinned by quartz glass, extremely resilient allows be left in situ indefinitely. hardware software of Silica have been co-designed...
We propose a new method, Speckle Flow SIM, to super-resolve every frame of dynamic scene from sequence diffraction-limited images at different acquisition timepoints. SIM uses fixed speckle-structured illumination encode the super-resolved information into measurements and neural space-time model exploit temporal redundancy for reconstruction. built simple, inexpensive setup experimentally demonstrated 1.88× super-resolution scene.
We demonstrate a single-shot miniature 3D computational microscope with an optimized phase encoder. Our method uses sparsity-based reconstruction to achieve 2.76-m lateral and 15،nm axial resolution across most of the 900 x 700 390،nm 3 volume.
Structured illumination microscopy (SIM) reconstructs a super-resolved image from multiple raw images captured with different patterns; hence, acquisition speed is limited, making it unsuitable for dynamic scenes. We propose new method, Speckle Flow SIM, that uses static patterned moving samples and models the sample motion during data capture in order to reconstruct scene super-resolution. SIM relies on sequence of images. The spatio-temporal relationship modeled using neural space-time...
We present an algorithm for improving reconstruction quality in diffuser-based single-shot 3D microscopy (Fourier Diffuser-Scope). decompose the raw video into temporal sparse and low-rank components, demonstrating on freely-moving hydra at 25fps.
We train space-time neural representations in an unsupervised, physics-based pipeline to reconstruct videos from DiffuserCam lensless imagers. Our method takes advantage of redundant information between frames, demonstrating improved quality when compared frame-by-frame approaches.
We couple a differentiable forward model of diffuser-based single-shot 3D microscopy (Fourier DiffuserScope) with an invertible and reconstruction algorithm to jointly optimize the imaging system parameters in memory-efficient manner.