I am not sure if the question is off topic but I need some input before I can finalize anything.

Actually to do my higher studies (in research) I wanted to make sure that I land my career in Digital Image / Signal Processing only. I have a strong background in Computer Application development. Now the Signal Processing course comes under Electronics And Communication Engineering / domain while Computer Vision & Speech Recognition subjects comes under Computer Science.

I am not sure if these subjects are different or doing Computer Vision / Speech Recognition is nothing but DSP / DIP. Is MATLAB being used for Computer Vision / Speech Recognition software?

Should I opt for Electronics And Communication Engineering course or one from Computer Science? Kindly guide me appropriately


1 Answer 1


These fall under signal processing and mostly you will find electrical engineers who are working in domain of speech recognition. The thing with computer vision is that it overlaps between both field.But as you mentioned you only have option of Electronics/Communications, Electronics is not much related to these field ,well in communications, we all use statistical signal processing which is also useful in speech processing and computer vision.

If you can get into Electrical Engineering(signal processing) that would be much better but out of these two I would suggest go for Computer Science if you plan a future in these field.

  • $\begingroup$ Thanks I would opt for Computer Vision / Speech Recognition in Computer Science stream. But one thing that I am not clear why DSP or Signal Processing is in Electronics Communication Engineering while the other is in Computer Science. In institute on CS people work on Computer Vision / Speech Recognition as it comes under their vertical. CS people do not cover Signal Processing as what electrical engineers do. So I assume knowledge of Signal Processing is not must to a greater extent? What are the areas each domain cover? What software CS and ECE engineers use? $\endgroup$
    – Programmer
    Commented Oct 11, 2014 at 3:15
  • $\begingroup$ The reason is mostly historical. DSP has historically been implemented as ASIC or VLSI, because the hardwiring gives the necessary performance for real-time DSP operation. Thus, it was in the realm of EE engineers. Only in the later part of last decade have general purpose Application Processors become powerful enough to be able to do DSP computations in realtime, on such processors. Many CS undergrad students take DSP as elective, and many CS grads take specialized DSP courses as part of their masters. The lines are rapidly blurring, but ... $\endgroup$
    – bdutta74
    Commented Oct 11, 2014 at 14:53
  • $\begingroup$ ... I am yet to come across EE grads who didn't have a semester length course on DSP. Also, note that many general purpose CPU's (Application Processors) having introduced specialized instructions to support DSP computation, often still fall short of compute power for newer advanced, complex algorithms, and often you'd find specialized DSP coprocessors in many modern gadgets, devices, computers -- take H.264 MP/HP video encoding for instance. $\endgroup$
    – bdutta74
    Commented Oct 11, 2014 at 14:53
  • $\begingroup$ Note that, to appreciate and understand the nuances of Computer Vision, you need a fairly good understanding of Signal Processing. However, there are many good engineers from CS background who have extensive skills in writing Computer Vision programs, thanks to high-level abstractions available via libraries like OpenCV. $\endgroup$
    – bdutta74
    Commented Oct 11, 2014 at 14:58
  • $\begingroup$ I completely agree with Icarus $\endgroup$
    – Omer
    Commented Oct 15, 2014 at 16:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.