"the use of computer vision techniques" - SOP for PHD in Computer Science

Hello everybody - I stumbled across this forum and it's a great source of inspiration for this kind of thing. If you could take a look at my essay I would appreciate it very much.

*****

Each generation of computer hardware allows the creation of more accurately rendered or visually interesting graphics. To achieve this increase in fidelity more time is required to produce a given piece of content; in turn increasing the cost of development. Procedural methods attempt to define the generation of content algorithmically, a large amount of output data can thus be created from a small set of input data. Using a purely procedural approach the quality of the content produced depends entirely on the choice of algorithm, a suitable solution may not be possible or be cost prohibitive.

I am interested the use of computer vision techniques to derive input data which could then form the basis of such procedural methods. The method of data extraction would be contingent on the phenomena of interest but both natural and man-made objects could make valid subjects. Previous work has focused on the creation of virtual foliage and the modelling of building facades, it would be my intention to extend this work into other areas. I would like to work with Dr **** who has experience in this domain.

I would like to pursue a career in research. I found the dissertations undertaken for both a Masters degree in Computer Science and Bachelors degree in Physics to be the most enjoyable aspects of each course. In my career as a software developer I find the most satisfying area of my work is creating a novel solution for a difficult problem.

My interest in the area outlined above stems from my MSc. dissertation entitled 'Procedural Generation of Virtual Cityscapes'. The motivation for the work was to create a system which would allow a user to create a 3D urban model with little effort. The system generated the geometry and texture of city buildings using a stochastic process with some weighting arising from user input. On completion the user could then perform a real-time fly-through of the city or export the model. I found this project intellectually stimulating. I liked the freedom to carry out research into related techniques and the creation of novel algorithms of my own. I also got satisfaction from implementing the software and viewing the end result. Previous to this I gained an interest in 3D graphics through games, in general I like creating things with a visual element, I am a keen photographer and have experience of graphic design earned through my job.

I think I have the skills necessary to carry out successful research in this field. A degree in Physics has provided my with skills in mathematics, I achieved my best individual results in subjects with a large mathematical component (Electromagnetism and Quantum Mechanics). I have developed programming skills through a career as a software developer, I have extensive experience of the design and implementation of suitable abstractions. I believe my research interests would be compatible with those of the [name of research group] and my work would benefit the department.

*****

Awesome, Hector, I hope you write an essay sometime about these subjects. I am very interested in the work of Robert Becker with bioelectromagnetism, which has implications for acupuncture. Maybe that is a little different from what interests you.

Anyway, check out The Body Electric by Becker. Maybe you will like it.

This essay really is strong enough to meet high standards. Mark caught some things that I had missed, and I'm glad he saw what I did not notice.

If anything, I would say you should make the writing LESS complicated.

Like... The method of data extraction would be contingent on the phenomena of interest but both natural and man-made objects could make valid subjects. -- very complicated, but really that is okay. You are complex in your thinking.