Photonic Interaction Lab

Our lab aims to create next-generation optical human-computer interfaces by integrating three synergistic domains: Nanophotonics, Optical Systems Engineering, and Physics-informed Machine Learning. We are fundamentally reshaping optical hardware to create new ways to perceive, interact with, and compute with the world using light. Conventional optical elements such as lenses and prisms are constrained by the natural properties of their materials, limiting device functionality and compactness. We break these limits using nanophotonics. By engineering artificial structures at the subwavelength nanoscale, we create ultrathin metamaterials and metasurfaces that allow precise control of the wavefront, intensity, polarization, and spectrum of light. While nanophotonics offers incredible control at the device level, its true potential is realized only when integrated into functional systems. A major focus of our lab is overcoming the fundamental trade-offs between form factor and performance in modern wearable displays. Real-world optical systems inevitably face physical limitations, optical aberrations, and nanofabrication imperfections that prevent them from achieving theoretical perfection. To bridge the gap between simulation and reality, we integrate computational optics and physics-informed AI into the hardware loop. By developing AI models grounded in physical principles and trained on actual experimental data, we can computationally predict, simulate, and invert complex optical processes. For instance, we use these models to computationally "pre-distort" input holograms, canceling out hardware errors and transforming a noisy prototype into a high-fidelity display. This hardware-software co-optimization extends across diverse platforms, including chip-scale quantitative