Since 1991, when I came to Pitt,
I went through several research and graduate projects working
signal and image processing,
nonlinear system identification and modeling,
fuzzy sets and systems.
Applications were primarily in the
field. I consider it a challenge to work on signals and systems
that are real, possibly coming from the living organisms.
Period prior to 1991 is covered in my
In our work at LCN, one
cannot adjust the parameters of simulation or the project requirements to
suit the imperfections of fancy theoretical results.
One has to deal with reality.
And reality is harsh! Signals are corrupted with noise which is anything
but Gaussian. The amount
of data is very large,
e.g. a typical EEG recording is 9 hours long, sampled
at 128Hz, resulting in 4.1 million of 12-bit samples per channel per
night per subject. These signals have to be preprocessed,
transformed, and analyzed automatically, focusing only at the most relevant
aspects of interest.
So, what does one do in such a situation?
Well, .....you may find interesting to read
what did we do about it.
The following are some of the topics, results and projects I am currently
working on. My intention was to provide simple explanations to
theoretical concepts behind my work in order
to bring you the basic ideas and motivations.
Bogdan R. Kosanovic