Research · I³ Group @ KAIST EE

Our Group's Core Technology

Computational Microscopy

We develop computational microscopy for non-destructive, high-resolution 3D imaging. By co-designing optics, acquisition, and reconstruction, we recover structure from measurements encoded through tomography, laminography, ptychography, and phase modulation. The 3D structure is not directly available from these raw measurements and must be inferred through inverse problem.

Low photon counts, limited angular/spatial coverage, aberrations, and scattering make this inference challenging. We integrate physical image-formation models with AI to estimate structure and imaging parameters. We aim to recover physically consistent structures while making the inference process more interpretable and the model itself more explainable.

Micro- to nanoscale imaging
Coherent and fluorescence imaging
Optical and X-ray modalities

Application 1 | Bio-imaging

Adaptive Optical Fluorescence Microscopy

We combine fluorescence microscopy, adaptive optics, and physics-informed AI for rapid and accurate aberration estimation and correction. We jointly infer sample's structure and wavefront aberration, aiming to improve imaging under low signal-to-noise ratios and large sample-induced aberrations while reducing dependence on wavefront-sensing hardware.

Research topics

Computational & physical adaptive optics Neural-field-based aberration estimation & correction Volumetric microscopy Joint sample–aberration inference Physics-informed reconstruction

Computational adaptive optics for microscopy — Kang et al., *Nature Methods* (2026)

Selected publications

Kang et al., “Coordinate-based neural representations for computational adaptive optics in widefield microscopy,” Nature Machine Intelligence (2024)
Kang et al., “Adaptive optical correction for in vivo two-photon fluorescence microscopy with neural fields,” Nature Methods (2026)

Application 2 | Metrology & Inspection

Computational Optical & X-ray Metrology

We develop optical and X-ray phase imaging for non-destructive metrology and inspection. Our physics-informed methods target reliable recovery of 2D phase and high-resolution 3D structure under photon-limited and limited-angle conditions, including deliberately reduced acquisition for largely accelerated inspection.

Research topics

Computational X-ray tomography & ptychography Sparse-view 3D reconstruction Photon-limited imaging AI-assisted reconstruction Optical-to-X-ray translation

Computational X-ray imaging for semiconductor metrology — Kang et al., *Optica* 10, 1000–1008 (2023)

Selected publications

Kang et al., “Dynamical machine learning volumetric reconstruction of objects’ interiors from limited angular views,” Light: Science & Applications (2021)
Wu and Kang et al., “Three-dimensional nanoscale reduced-angle ptycho-tomographic imaging with deep learning (RAPID),” eLight (2023)
Kang et al., “Attentional Ptycho-Tomography (APT) for three-dimensional nanoscale X-ray imaging with minimal data acquisition and computation time,” Light: Science & Applications (2023)
Kang et al., “Accelerated deep self-supervised ptycho-laminography for three-dimensional nanoscale imaging of integrated circuits,” Optica (2023)