After m.s in computer vision, should i go for industry or phd

[mish4]

So recently i have been thinking about what to do after i graduate with my masters of science degree. I did my undergrad in electrical engineering at Penn State University, and stayed for a masters degree.

My original motivation to get a masters was to take more classes in computer vision (feed my interest in the subject) and to see how i like research (never did research in undergrad).

Well after completing one year of my degree, i have taken several classes (i learned a good amount), and began doing research in medical imaging. I feel like its kind of early for me to tell if i like research or not. Sometimes research feels slow, or even feels like you are working on a lot of implementation details and not so much the core problem. Some days its exciting, other days not so much.

Come this fall i have the opportunity to apply for phd programs for the next fall, which by then i will have graduated. However I don't know whether I would want to spend another 2-3+ years in school. I do feel like I am always unsatisfied with how much i know, and thus want to take more classes to fill in some gaps.

Recently I have been thinking since a lot of computer vision is based on algorithms, i should try and catch up in that area. Being that my undergrad was ee, i didn't get the chance to take discrete math, and algorithms courses. I have started reading some stuff on discrete math, and plan on taking an algorithms class in the fall.

So one reason i would want to go for a phd is to take more classes and learn more. However the real emphasis of a phd is on the research project. At this stage i really don't know if i would enjoy spending 2 years on one project. There seem to be so many factors involved (advisor, the project, whether you can keep motivated, candidacy exams). After all a phd is a serious time commitment.

The other option is to go into industry / or research lab and try and find a challenging computer vision job that would want a masters student (although i don't know how many research labs want masters students).

From my browsing the web it seems like a lot of computer vision jobs are for phd students.

I know that I want interesting work that will be challenging (and related to vision / image processing), but i don't know if that is only given to phds in the computer vision field.

any word of wisdom are appreciated


Thanks,
Mikhail

 

[s1cko]

So recently i have been thinking about what to do after i graduate with my masters of science degree. I did my undergrad in electrical engineering at Penn State University, and stayed for a masters degree.

Thats exactly where I want to go to school! I have questions for you if you don't mind. Sorry I just quoted part of it and didn't offer any real advice.

 

[Choppy]

It seems to me like you're aware of the issues at hand. If research is where you want to go, a PhD is the degree to have. Ultimately it comes down to whether or not you want to make that kind of committment. What might help in your decision is to investigate some projects that you find interesting. As you've mentioned, research is not always about working on the core problem. Some days you can spend an entire afternoon just figuring out how to format the graphs you want to include in your next paper. But I think all jobs have their mundane tasks, and you won't be able to get away from those.

 

[kote]

It seems to me like you're aware of the issues at hand. If research is where you want to go, a PhD is the degree to have. Ultimately it comes down to whether or not you want to make that kind of committment. What might help in your decision is to investigate some projects that you find interesting. As you've mentioned, research is not always about working on the core problem. Some days you can spend an entire afternoon just figuring out how to format the graphs you want to include in your next paper. But I think all jobs have their mundane tasks, and you won't be able to get away from those.

You may also want to consider the job market. Would you actually be able to get a satisfying industry job right now? I'm not implying that it's impossible, but it's hard at the moment. Then again applications to grad school will also be up. If nothing else, it can't hurt to apply to both.

 

[comp_math]

Do a PhD. Try to get into top schools like MIT, Caltech, Stanford, UCB, and CMU. Good luck.

 

[mish4]

So I have decided to apply to phd programs for computer vision. I know i enjoy working on computer vision / image processing projects. I also know i want to learn more about computer vision.

The thing I am finding hardest, is writing the statement of purpose. Specifically writing about specific programs. I spent about 2 hours looking at all the faculty and research going on at cmu. I don't know how to tie that into my statement of purpose. (each school i apply to will require me to write why i think they are the best school for me, ie their research groups.)

I guess i just find it really hard to say "i would love to work with professor so and so, because i truly find his research inspiring and interesting" Or even "this research lab seems like a perfect fit for me because they do research in polygon convex optimization for scene understanding" ..

the reality is, research topics are complicated and its hard to get a good sense of the research without spending a few days reading a paper and trying to understand it.. i know my general research interest within computer vision (tracking/registration/reconstruction/image processing) but going a lot more specific than that... i can't do.

anyone have any ideas how i should go about this part of the statement of purpose..

thanks a lot!