An overview of the issue

✍ Scribed by John Hershey; Rao Yarlagadda

• Publisher: Elsevier Science
• Year: 1992
• Tongue: English
• Weight: 83 KB
• Volume: 2
• Category: Article
• ISSN: 1051-2004
• DOI: 10.1016/1051-2004(92)90024-s

✦ Synopsis

An Overview of the Issue

Signal processing is an exciting field, as I've said many times. I have often thought it would be fun to write a short piece on the differences and similarities between signal processing and statistics. Aside from the fact that signal processing folks usually have a sense of humor, there are some very interesting counterpoints. Anyway, it is, as we've also often said, in the merging of disciplines that we make great leaps. Our first article is a true synergy. When I first ran into the work of these researchers, I was quite impressed and kept after Badri to do a piece for our journal. He and his colleagues did a fine job, I think. They were able to impose Boolean and other syntactic constraints on detection and estimation problems. It is an unusual but extremely important marriage. I expect that their paper will motivate a lot of thought and further fruitful research.

Nasir Ahmed and Bruce Armstrong follow with a very nice observation concerning what may be an important invariant behavior respecting the DCT. It is important that we continue to study this particular transform as it gains utility in the discipline and standards of image compression.

Continuing on, we have two papers which will be of interest to those concerned with speech representation and recognition. This is followed by a paper by Knittle and Magotra that discusses a technique that must cope with a very difficult "real-world" problem -that of enhancing very short duration signals in severe clutter.