Software & Datasets

Imaging Through Aberrations

Neural Wavefront Shaping (NeuWS)

Code and data demonstrating neural wavefront shaping for imaging through static and dynamic scattering media.

Guidestar-Free Adaptive Optics with Asymmetric Apertures

Code for guidestar-free, sensorless closed-loop adaptive optics using asymmetric apertures and machine learning.

SLURPP: Single-Step Latent Diffusion for Underwater Image Restoration

Code for single-step latent-diffusion underwater image restoration, plus the physics-based synthetic underwater image-generation pipeline.

AONeuS: Acoustic-Optical Neural Surface Reconstruction

Code for a physics-based framework that fuses optical and imaging-sonar measurements for underwater 3D surface reconstruction.

Transmission Matrix Dataset

A public dataset of optical transmission matrices for testing phase retrieval and imaging-through-scattering algorithms.

Non-Line-of-Sight Imaging

Keyhole Imaging

Code and data demonstrating non-line-of-sight imaging and tracking of moving objects along a single optical path.

Seeing the World through Your Eyes

Code for reconstructing the scene around a person from the reflections captured in their eyes.

Deep-Inverse Correlography

Neural networks for imaging around corners using phase retrieval.

HDR Imaging

Deep Optics for Single-Shot HDR

Code and data demonstrating single-shot high-dynamic-range imaging with learned optics.

Depth Sensing

CodedVO: Coded Visual Odometry

Code and dataset for metric-scale monocular visual odometry using optically encoded depth cues.

Depth from Defocus with Learned Optics

Code and data demonstrating monocular depth estimation with learned optics.

Compressive Sensing & Phase Retrieval

FPM-INR: Fourier Ptychographic Microscopy with Implicit Neural Representations

Code and data for reconstructing Fourier ptychographic microscopy image stacks with implicit neural representations.

Learned D-AMP, D-AMP & D-prGAMP Toolbox

Neural networks and algorithms for compressive sensing and compressive phase retrieval, including code to train with SURE loss instead of MSE.

Deep Simultaneous Source Separation and Phase Retrieval

Code to solve simultaneous source separation and phase retrieval (S3PR) using deep generative models.

prDeep

A neural-network-based, noise-robust phase retrieval algorithm.

D-VDAMP

Variable-density sampled compressive MRI reconstruction code.

Event-Based Imaging

Repurposing Video Diffusion Models for Event-based Video Interpolation

Code for adapting pre-trained video diffusion models to event-based video frame interpolation.

Reflection Separation

Flash-Split: 2D Reflection Removal with Flash Cues and Latent Diffusion

Code and data for separating transmitted and reflected light from a (potentially misaligned) flash/no-flash image pair using latent diffusion.

Flash-Splat: 3D Reflection Removal with Flash Cues and Gaussian Splats

Code for separating transmitted and reflected light in 3D using flash/no-flash cues and Gaussian splatting.

Uncertainty Estimation

Hyper-Diffusion (HyperDM)

Code for estimating epistemic and aleatoric uncertainty with a single model, matching the accuracy of large diffusion-model ensembles.