Personal Statement Of Purpose/UCSD/Computer Vision
Oct 27, 2016 1508 Words 7 PagesView in Writing Tool
Statement of Purpose / UCSD / Computer Vision / Ph. D.
Sung-Pil Paek
This statement applies for admission to the Ph.D. program in Department of Computer Science and Engineering of UCSD, specifically, to continue my study in computer vision as a subfield. I indeed am interested in recognizing, identifying, and detecting the real world through computers in particular with visual information. Based on an understanding of our surroundings of life, I always enjoy replacing the time-consuming, hectic works with computer software and let it be a better choice than other routine works. Hence, I am fascinated to deal with automatic annotations and tracking of images or videos, which are my current research interests.
When I was a student in College of Information & Communication of Korea University, one of most competitive institutes in South Korea, I have taken series of lectures from Professor Jung-Hyun Han. His research areas include physics-based computer simulation, and he taught “Game Programming”, and “Introduction to Interactive Contents” as a course at the college. Through attending his lectures, I was favored by learning about computer graphics, including concept of applied mathematics, applied physics, and projection of three-dimensional objects into two-dimensional space. I further recognized that reconstruction of the three-dimensional objects from a photograph can be an exciting topic and also learned that this was a part of the computer vision area. Such personal academic experiences in the undergraduate level stimulated me enough to seriously start deliberation in the field which eventually inspired me to be a professional researcher in the future.
In fact, my preference in continuing the study for Ph.D. degree was originated from an opportunity of taking part in a research when I was an undergraduate student in Korea University. In January 2014, I got involved in a study of different fields conducted in Biosensor System Engineering Laboratory with the help of Professor Se-Hwan Paek in the same university. The research group was developing a plastic-based biosensor which measured the concentration of a protein, indicating myocardial disease and displayed the outcome by color as a signal. My task given in the study was to build a mobile detection system that captures the signal as an image and then quantifies the image by graphing the optical density against the position. It was very troublesome at the initial time period because I didn 't even know how to build an empty Android application. For all the trials and errors, I enjoyed discussing with the lab members with whole different knowledge from that of mine, which had greatly benefited the development process. Six months after the onset, I finally had photographed the magnified image of the sensor, aggregated all RGB data per vertical pixels, and transformed them into a quantified graph exported to Excel. This work meant significant because an expensive machinery should be used unless such attempt was made. Furthermore, I was convinced to believe that the mobile detector may be possible to conduct a medical diagnosis without being constrained by time or location. My work eventually got included in a publication, and I was honored as a co-author of the paper published in a journal, "Biosensors and Bioelectronics", ranked in a high impact (impact factor: 7.476).
Further, I determined to personally continue the study such that the mobile detection system can automatically search the output signal, while this process was currently carried out using human eyes. After looking around for any possible solutions, a technique based on ‘machine learning’ which was taken as the course in the university was considered as a candidate means for the automation. Obsessed with solving the problem, I attempted to use the technique, although knowledge regarding the fundamentals was not sufficient. The naïve bayes classifier was tried to automate the sensor, but training sets or knowledge were not enough to see meaningful effects. Although I wasn’t able to deal with the issue in the past, I hope to dig out a solution if opportunities are available, which motivated me to pursue more studies in computer vision area.
In order to gain deeper insight into the programming skills and computer science, I had decided to work in an IT company, Ivy Solution Co., before graduation of the university. It should be noted that this work was the substitute of mandatory national service in my country. Hence, I was employed as IT Professional and developed web-based portal system for three years. I was trained how to practically develop the system with JSP and maintain Linux Servers as well as DBMS. After completion of the national service in 2013, I graduated the university and spent another year working in an industrial database research lab, TmaxData Co., which is a renowned middleware developer in Korea, and then to work in a university, Yonsei University. In the company, I developed and maintained mainly client-side program, Open Database Connectivity (ODBC), which is written in C, and re-factored existing database installer written in Java. As a skilled researcher, I also had several opportunities of presentation in front of CTO, Dr. Dae-Yeon Park, of the company. In the presentation, I demonstrated how to maximize compatibility between the company’s database, “Tibero”, and Oracle’s client-side products to attract prospective customers. Lastly, to experience various types of work, I moved to Academic Information System Team of Yonsei University. In the university, I developed and maintained university portal system, mostly written in JSP and HTML5, for students, faculties, and other personnel. Through total four years of job experiences, I gained not only diverse programming skills, but also became confident about undertaking unknown issues without mental contraction.
Along with the definitive reasons to continue my study, it is my belief that the ability to communicate with people is also an integral part of being researcher. So, I should introduce an extracurricular activity, which also gave an effect on me up to now, as college’s student ambassador named ‘Yeoul’. Not only was this one of the most precious participation in my life, it also pushed me to hone fundamental aptitudes. In the activity, I was likely to be a teacher because I needed to present information in ways others easily understood. As a representative of our decent college, I needed to know more about our school than anyone else, and I needed to explain the facts more easily than anyone else. Furthermore, I naturally met many people and various types of people from elementary school students to foreign university presidents. By meeting so many people, I was able to overcome my introverted personal traits and let myself socialize with my colleagues. Indeed, I have become confident and comfortable in expressing whatever I have in my mind. Above all, I had devoted myself to the job over total six hundred hours, which I am proud of.
Among graduate schools in the United States, there are good reasons that UCSD is the right place for me to continue my study for Ph.D. in computer vision area. UCSD supports the computer vision research with tremendous resources and attention by launching the Center for Visual Computing in April 2015. Since the center is also sponsored from well-known media and tech companies, there is no doubt that it has the potential for major societal impact. Among many research labs in the center, I would like to work particularly with Professor David Kriegman. His researches on computer vision indeed apply to many areas such as social networks (urban tribes) and coral ecology (benthic survey). In the latter, I have enjoyed reading a recent paper about automatic annotations of coral reefs from large-scale benthic survey images. The paper reported that reflectance images itself won’t always contain sufficient information for proper classification of coral reefs, and further suggested to combine those reflectance images with fluorescence images to obtain more accurate results. This discovery is spectacular for its originality, utility, and extendibility relatively to other ecological applications. As my intense interest in computer vision was introduced earlier, I am confident that I can contribute to Professor Kriegman’s diverse researches in attaining achievements during my Ph.D. study if the chance is provided. Upon graduation, I am eager to be a professional researcher dedicating the remained life to research in the same field.
I always follow my zest for enthusiastic wish to grasp the happiness I felt when I was getting closer to understanding of the problem or something I wanted to know. Working experiences in the industry where the type of solution was its secondary concern wasn’t enough to fulfill my insatiable need. More importantly, I want to make an impact and be the authority on my academic field by doing it right, not relying on the heuristics, even though it may take a time. This is what I have realized over the past years and, after all, the joy of research is what I have been searching for myself desperately. I am appreciative of the chance to transfer my unique story to the admission committee, and I express my gratitude towards your time and effort. Thank you.
页:
[1]