Prof. Edmund Lam delivered an invited talk at CIOP 2025 on advanced deep learning in digital holography, highlighting ISL's work in end-to-end learning and physics-integrated networks.
Our differentiable imaging framework handles misalignments, aberrations, and low-quality data in a unified, fully differentiable pipeline for Fourier ptychography.
A comprehensive review of differentiable imaging, outlining progress, challenges, and the outlook toward digital-twin-equipped computational imaging systems.
CUBE generates controllable videos from event-camera edge streams guided by text prompts using pre-trained diffusion models—no additional training required.
Prof. Edmund Lam delivered an invited talk at CIOP 2025 on advanced deep learning in digital holography, highlighting ISL's work in end-to-end learning and physics-integrated networks.
Our differentiable imaging framework handles misalignments, aberrations, and low-quality data in a unified, fully differentiable pipeline for Fourier ptychography.
A comprehensive review of differentiable imaging, outlining progress, challenges, and the outlook toward digital-twin-equipped computational imaging systems.
CUBE generates controllable videos from event-camera edge streams guided by text prompts using pre-trained diffusion models—no additional training required.
Plenary talk at Optica Imaging Congress on event-sensor-based computational neuromorphic imaging and the evolution of imaging technologies.
