The advent and development of photonics in recent years has ushered in a revolutionary means to manipulate the behavior of light on the nanoscale. The design of photonic structures and devices, to date, has relied on the expertise of an optical scientist to guide a progression of electromagnetic simulations that iteratively solve Maxwell's equations until a locally optimized solution can be attained. Innovative approaches and tools play an important role in shaping design, characterization and optimization for the field of photonics. As a subset of machine learning that learns multilevel abstraction of data using hierarchically structured layers, deep learning offers an efficient means to design photonic structures, spawning data-driven approaches complementary to conventional physics- and rule-based methods. The objective of this PhD thesis is to explore deep learning models and optimization approaches for the design of future photonic devices, with various applications such as imaging, hologram, sensing, and display. In specific, the theme of thesis is to utilize various deep generative models to find simple representations for highly complex photonic structures, such that optional optimization algorithms can be efficiently applied to identify the photonic structures with optimal performance. The developed design framework has potential applications in the optimization of future highly compact optical systems such as photonic computing, LIDAR, telecommunications, and virtual/augmented reality display.
Cite this work
Researchers should cite this work as follows:
MATIN Development Team