Embedded Adaptive Control/Signal Processing Industry Careers

Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
1 reply · 2K views
MassivelyParallel
Messages
1
Reaction score
0
Hello,

I just began a PhD in Computer Engineering three weeks ago. I have an undergraduate degree in EE from a lesser quality university than I am now attending.

My background in EE focused on DSP/Image Processing, but it was all deterministic stuff (so no real-world stochastic stuff besides basic random signal theory).

I now get to determine how to focus my PhD and can basically go in any direction I want. My PhD advisor is a computer architecture specialist, and not a specialist in signal processing or control systems. He mainly hired me to be the "signal processing guy" to work on research focusing on analyzing how very specialized image processing algorithms utilize hardware resources of parallel processor architectures (these algorithms were not developed with parallelism in mind).

Well, I have been very fascinated with both signal processing and control systems, and was wondering if I should try to focus these years of my PhD on becoming a specialist in the area of applying adaptive signal processing and control systems algorithms in embedded hardware.

I am also interested in computer vision and machine learning, and these area's seem like version 2.0 of image processing and adaptive signal processing.

I definitely want to work in industry after my PhD.

Does anyone here work in industry in a field related to embedded signal processing and control (or computer vision and machine learning)?

Is this a hot field?
 
Last edited:
Physics news on Phys.org
Well the first question is: does this fire you up? Do you think you could be passionate about this project? You will work harder than you ever thought possible on it, so you had better enjoy it.

That said, custom processor architectures for specific applications or software domains (e.g. image processing) is a very hot area, and will most likely be hotter when you get out. A lot of researchers in high-performance and warehouse-scale computing are finding that they are power limited, and there is a potential for significant efficiency improvements if they move to processing architectures co-designed with their software architecture.

So. my feeling is yes, this is a hot area, and if you are excited about the work, go for it!