原标题：【PS范文】— 非大牛录取卡耐基梅陇(CMU)计算机硕士的PS分享

申请人背景

北京某高校，计算机专业， GPA：3.3，GRE：315+3，TOFEL：92

录取结果

(1) 卡内基梅隆大学 Master of Science in Computer Science (该专业近年毕业后年薪100K+美元 )

(2) 加州大学欧文分校 Master of Science in Computer Engineering

个人陈述原文

申请目标：美国顶级计算机硕士/博士专业

My long term exposure to research training has made computational thinking my conditioned reflex. When catching sight of birds flocking together on the grassland, my first reaction was that it can be used for feature clustering, a topic on which I began a research project two years ago. As I gather more research experience, my interest in computer science has grown substantially. While I sometimes feel frustrated and foolish, unable to solve a problem, I am often satisfied when making a breakthrough, like an “Architect” constructing space-time. In the journey of research, simultaneously fraught with frustrations and joy, the only thing becoming increasingly clear is my passion for using technology to change people’s lives.

In order to improve my research capability and expand my professional outlook, I have participated in several different research projects while an undergraduate at XX University. While an intern at Baosight Company, I was inspired to join one of their ongoing projects setting up video surveillance that monitors a large square. Currently the surveillance requires human supervision, which is quite time-consuming and prone to human errors. I developed an automatic approach that could recognize crowd behavior computationally efficiently and effectively, removing humans from the equation. I used Webber’s law, which is frequently used in image analysis, and applied it to videos of crowd motion. Experimental results validated that our proposed algorithm outperforms the current algorithms and can segregate abnormal behavior from normal crowd behavior effectively.

While also at XX University, I joined another project team working on cell segmentation imaging, aimed at using interactive segmentation to improve resolution accuracy and distinguish image features computationally. In the project, I was responsible for constructing a discriminative grid, assigning small distances between like, intra-class pixel samples, and maximizes distances of inter-class samples. The algorithm has proved to be effective for semi-supervised learning based on label propagation. I tested the algorithm interactively, marking cell types by hand as initial conditions and then compared to two automatic segmentation techniques, namely Interactive Segmentation via Random Correction and Segmentation without Interactive Correction, using Matlab. The experimental results showed that our proposed algorithm converges to desirable results rapidly and requires few human interventions.

My research interests are wide-ranging. In my sophomore year I attended a lecture delivered by Dr. Zhang Yaqin from Microsoft Research Asia, who shed light on opportunities to apply camera video processing to gaming, which spurred my intense curiosity in related technologies. This led me to participate in a SRT (Students Research Training) program entitled video semantic sketch extraction and stylization. In this program, I conducted sketch (i.e. line drawing) extraction work on color images with a group of graduate students. While we encountered difficulties with some images due to noise, we solved our problem by improving the Breadth-First Algorithm, detecting the connected region of two-dimensional sketch and filtering out unimportant lines. I found that sketch extraction is related to cognitive psychology. From the perspective of human cognition and sketch semantics, I extracted images’ salient regions, enabling me to accurately describe the foreground information of images and complete sketch extraction. Our extraction algorithm achieved better effects than previous technology and hence warranted a national patent in China.

To build up a solid knowledge in my major, I have eagerly taken all the related courses XX has offered. Putting my knowledge into practice was what inspired me most in all of my coursework. For example, at Professional Practice course, I used my improved continuous speech recognition algorithm to enact voice commands on the mail system of mobile phones, winning me the first prize in the course. When taking Fundamentals and Applications of Multimedia Technology, I implemented the color histogram based image retrieval by calculating the histogram of the query image and measuring the distance of histograms. These experiences make me feel a great sense of accomplishment, demonstrating my dedication to academics and research.

I am eager to start my PhD at University of California, Los Angeles (UCLA) because of its fantastic strengths in the area of computer science, particularly with the world-class VCLA lab and vision lab. When I arrive, I will quickly pursue projects under excellent professors and role models in this field, for example I could work with Dr. _______ ________ on _____[name type of research this person does]_____ or Dr. ______ _______ on _____[type of research other person does]_____.

I am confident that my diverse research experiences have prepared me for the exceptionally strong PhD program at UCLA. Currently, I’m working on research related to recommendations in social media and I keenly expect to continue breakthrough research during my PhD at UCLA. After having obtained my PhD, my goal is to return to a Chinese university as a research and teaching professor, disseminating cutting-edge knowledge and stimulating students’ creativity with my trained insight. In the meantime, I will continue to doing my research with my “conditioned reflex” and develop socially beneficial products, which constantly push the boundaries of innovation.

我们是谁？

Adam

Cornell University

Tyler

The University of New South Wales

University of Sydney

University of Western Australia

Vittorio

Yale University

Yale University

David

Claremont Graduate University

Pomona College

Nina

Stanford University

Yale University

责任编辑：