MIT︱PhD︱Personal Statement

Have you ever heard of Go? AlphaGo, an artificial intelligence specialized in playing Go, shocked the world with its exceptional performance in 2016. As one of the top-tier Go players in Taiwan, I believe it is the dramatic improvement in “global view” that leads to its breakthrough. “Global view” is the most valuable asset I have gained from playing Go, with which I can effectively utilize ample information and resources from different perspectives to achieve optimal results. That’s why I am intensely fascinated by Optics and Photonics, a highly interdisciplinary subject involving versatile research fields and applications. With that being said, Massachusetts Institute of Technology’s Department of Electrical Engineering and Computer Science, renowned for its interdisciplinarity, is an ideal place to pursue my Ph.D. degree.

To begin with, I have a solid background in Optics and Photonics. Receiving A+ in all major subjects, I graduated with the first place from the Graduate Institute of Photonics and Opto- electronics (GIPO), National Taiwan University (NTU). My course selection is systematic, starting from the fundamental theories of optics and electromagnetics (e.g., Fourier optics and optoelectronic electromagnetics), and extending to the applications and numerical methods for engineering problems (e.g., plasmonics and mathematical physics). Instead of focusing on knowledge acquisition merely, I have also put considerable effort into learning the engineering mindset, that is, utilizing the knowledge to solve real-world problems. For example, in the course Simulation of Light Scattering and Propagation, I incorporated the finite-difference method with the ray-tracing method to simulate the optical caustic effect of water waves. This work was highly appreciated by the professor and used as the teaching material in the following years.

In addition to my academic background, I have been engaged in four major research projects that help me to outline my research interests and goals. During my junior and senior years at the Department of Electrical Engineering, NTU (NTUEE), I launched my first research project regarding objective lens design using microlens arrays. I adopted ray transfer matrix analysis with Zemax OpticStudio and introduced the genetic algorithm for optimization, eventually coming up with an objective lens design with 5-time shrinkage of size, which is beneficial for the promotion of home care. This work received undergraduate research funding from Ministry of Science and Technology (MOST), Taiwan, and I was also invited to give an oral presentation of my research to the committee, an opportunity that only five among hundreds ofcompetitors were offered. In this project, I have learned the design rules and techniques for optical system design.

After entering graduate school, I carried out the second research project, my master thesis, involving super-resolution imaging characteristics of microsphere microscopy. The typical size of microspheres for super-resolution microscopy ranges from 2 to 50 um, a mesoscopic scale which lacks efficient but precise simulation approaches. In particular, the complexity ofnear-field imaging and the virtual-image formation of microspheres hamper the theoretical and numerical analysis for predicting the experimental observations. After scrutinizing various techniques in computational electromagnetics, I integrated FDTD, near-to-far-field transformation, and angular spectrum method to establish a full-wave simulation architecture capable of predicting the magnification and effective resolution with an error less than 2%. On top of that, the sample- dependent resolution has been demonstrated for the first time in the simulation. This break- through could facilitate the development of microsphere microscopy in practical applications such as biomedical imaging. The relevant study has received “Student Paper Award” in Optics and Photonics Taiwan, International Conference (OPTIC) 2018, and this work is currently submitted to OSA Journal Optica for publication. More importantly, I achieved considerable proficiency in the software tools for electromagnetic modeling including Lumerical-FDTD, MEEP, and MPB.

During the last year of my graduate study, I took part in the collaborative research mutually conducted by NTU and Applied Materials, Inc. The goal was to bring out a system-level architecture of autostereoscopic 3D displays composed of high-directionality OLEDs and metasurfaces for light deflection. Inspired by the last two projects, I carried out the two-stage optical simulation strategy: FDTD for calculating the emissive characteristics of OLEDs and the phase modulation of metasurfaces, and the ray-tracing method for analyzing the change of far- field emission profile in between. This approach mitigated the burden of computational resources for simulating such a multiscale optical system. Finally, the research group successfully proposed architecture with comprehensive concerns of technical specifications and design trade-offs. This study is currently organized for patent application. Through this project, I have gained valuable experience in the system-level integration of optical devices.

Apart from the research at NTU, I joined Research Center for Information Technology Innovation (CITI), Academia Sinica for about one and a half years. At CITI, I embarked on another branch of my research in Computer Vision, involving semantic matching of images via deep learning and cycle-consistency. It was rather atypical for a student majoring in Optics and Photonics to get involved in such a different research field. From my own perspective, however, an ultimate optical system should consist of hardware design for data acquisition and software design for data interpretation to remedy the respective deficiencies. After learning the cutting- edge techniques in Computer Vision, I started keeping my eyes on its applications in super- resolution microscopy, AR/VR, and light-field displays. Thanks to this “atypical” experience, I am able to think out of the box, drawing connections between different disciplines.

In these research projects, I have progressively consolidated my research interests and goals. For now, my research interest primarily focuses on, but not limited to, advanced light steering techniques in nano-optics such as metamaterials and photonic crystals. Afterward, I aspire to move forward to multiscale optical system integration, bringing exceptional features in nano- optics into the system-level design and real-world contexts. My ultimate career goal is to closely incorporate optical systems with techniques in Computer Vision, realizing a system architecture with super-efficient acquisition and versatile utilization of information in the real world.

My passion and ambitions prompt me to seek high-quality training at MIT, which can enhance my professional knowledge and skills to fulfill my goals. MIT’s Department ofElectrical Engineering and Computer Science possesses rich resources of huge assistance for me, at which I could have access to chromatic research fields and, in particular, its prestigious faculty. I am fascinated by the research on photonic crystals and integrated optics conducted by Prof. Leslie A. Kolodziejski. Also, I am of great interest in Prof. Qing Hu’s research on THz imaging. Moreover, Prof. James G. Fujimoto’s research on OCT also draws my interest. I feel exceedingly excited about pursuing my Ph.D. degree at MIT, not merely because it helps me prepare for my future career, but because this process itself must be challenging and thus rewarding.