In this compelling episode, Dr. Weijian Yang introduces us to a breakthrough in computational imaging: a compact and learnable lensless 3D camera capable of real-time photorealistic imaging. This technological innovation holds great promise in revolutionizing how we capture and interpret our world.

The device, developed by Dr. Yang and his team, replaces traditional bulky optics with customized thin optical masks, making it lightweight and portable. Contrary to existing lensless imaging methods that require extensive calibration and heavy computational resources, this lensless 3D camera overcomes these challenges. Dr. Yang discusses the custom design and fabrication of the optical phase mask, optimized for spatial frequency support and axial resolving ability.

The team has developed a robust physics-aware deep learning model with an adversarial learning module that allows for real-time depth-resolved photorealistic reconstructions. The ability of this lensless imager to resolve depth and "see-through" opaque obstacles offers transformative potential across a range of computational imaging applications.

Join us as we delve into the future of imaging technology and its implications for a variety of fields, from medical imaging to remote sensing and beyond.

Key Words: Computational Imaging, Lensless 3D Camera, Real-Time Imaging, Photorealistic Reconstruction, Optical Phase Mask, Physics-Aware Deep Learning, Adversarial Learning, Depth Resolution, See-Through Imaging.

Tian F, Yang W. Learned lensless 3D camera. Opt Express. 2022 Sep 12;30(19):34479-34496. doi: 10.1364/OE.465933. PMID: 36242459; PMCID: PMC9576281.