• I would suggest that in the future more test problems be broken up into parts (a,b,c, etc). That way, it may be possible to demonstrate understanding of part of a topic and receive part of the credit. I understand that partial credit is given on long problems, but I feel that this credit would be more fairly distributed if certain parts of the problem were assigned different point values.--Jwromine 23:12, 23 April 2009 (UTC)
  • Since we did LPC in the lab without going over the details of solutions in class, I would prefer to learn the math behind the two methods before doing it in lab.
  • How about less review of ECE301 and ECE302 material so we can do more speech/sound/image processing instead? --Mboutin 16:08, 28 April 2009 (UTC)
  • I think you can definitely cut down on some of the review material. All of the CTFT, DTFT, and ZT stuff could probably have been grouped in one homework, maybe two if you incorporate sampling. I'd imagine all of that, plus DFT, FFT, and circular convolution could be covered by the first exam. Next maybe a short review of 302 material, then have the second exam be all speech processing and the third all image processing. With regards to the "quiz" we had on Wednesday, I enjoyed learning both speech and image processing and would hate to have to sacrifice one for the other. One thing to keep in mind when changing the structure of the class, however, is the labs. It's painful when a class and it's lab don't correlate, so if you change the material significantly the labs may become obsolete (unless, of course, you rewrite them). Other than that the more practical the material, the more interesting it is. --Pjcannon 11:20, 1 May 2009 (UTC)
    • This is excellent feedback Phil! Thanks. --Mboutin 10:25, 1 May 2009 (UTC)
  • I was a little disappointed by the amount of image processing we covered. As an individual who enjoys photo-editing, it was really neat for me to learn how most of the functions in the programs and tools I use to edit photos work. However, it seems that we only went over the basics in this area, and we did not cover it as in-depth as the other topics, mainly due to lack of time. If the review at the beginning was shortened, we would have more time to go over image processing material. With a processor whose specialty and main area of interest is image processing, it is a shame that we did not have more time to discuss this area of DSP. That's about the only thing I would suggest. Over all, I think the course went pretty smoothly this semester. --vhsieh 21:41, 1 May 2009 (UTC)


  • I would like to see more examples in class to go along with the materials. Especially those related to homework, otherwise maybe we should change the homework to go along with lectures. Sometimes I really feel the homework and lectures don't quite match and it takes a long time to go through notes by previous professors just to figure out how to do homework. --jgeng
    • I agree with doing more examples in class. I think I learn best by seeing examples, doing practice problems, and seeing applications. Going through theory is useful, but there should be a balance between theory and concrete examples in lecture. --Leedj 17:03, 3 May 2009 (UTC)
  • When I took ECE302 with Prof. Bell, we had "homework 0" assignment, which basically forced everyone to review the maths they needed for the course, but it also gave the professor some idea of which topics needed more review than others. I don't think it was counted for credit. This might be useful in future semesters, if you don't mind extra assignments. Also, concerning the balance of speech and image processing, I thought the distribution was pretty good, considering the amount of time available for both. I found speech processing to be more interesting than images, but both are somewhat necessary. I would insist that you don't remove the speech Lab Experiments at the very least; making a vocoder in software was the coolest thing we did all semester :) --Mreeder 17:42, 3 May 2009 (UTC)

Alumni Liaison

Ph.D. on Applied Mathematics in Aug 2007. Involved on applications of image super-resolution to electron microscopy

Francisco Blanco-Silva