On the Linkedin group dedicated to compressive sensing (the group is 696 members strong), there is this question on the availability of a code featured in several papers listed here.. The person who requested some details on how the algorithm works, eventually requested the code and wrote:
The first papers started in 2007 and some are still being generated in 2010. Obviously there is much work performed on the code and nobody requests that the source code be handed out to anybody. However, it is really a question of reproducibility. At the very least the authors should give outsiders the means to reproduce the figures of their website and papers. In the end, as Donoho said (or was it Claerbout ?), if authors cannot release a way to reproduce the figures of the paper, you really have to come to terms that these results are anecdotal and you should not spend much time and effort to reproduce them. As a student, I once faced the issue of reproducibility, only to find out that, to be charitable, I was right and the computational paper I was comparing my computations to, had, shall we say, suboptimal results. In retrospect, the only error I made was not write a letter to the editor on the matter, it would have been a reference at a minimum. Then again, what's the point, we're talking about a field that really has not evolved for the past twenty years. If you want to read more on my thoughts about why reproducible research should be your main priority, check Nobody Cares About You and Your Algorithm.
In a totally different direction now, I have been asked this question by a few of you on how to cite Nuit Blanche as a reference.. I am not sure this is such wise thing but since the requests are there, let me give it a try by following these guidelines:
"...About if I've tried to contact with the authors, but only one of them answered my e-mails and telling me that he can't help, because the code of that example can't be distributed,..."
The first papers started in 2007 and some are still being generated in 2010. Obviously there is much work performed on the code and nobody requests that the source code be handed out to anybody. However, it is really a question of reproducibility. At the very least the authors should give outsiders the means to reproduce the figures of their website and papers. In the end, as Donoho said (or was it Claerbout ?), if authors cannot release a way to reproduce the figures of the paper, you really have to come to terms that these results are anecdotal and you should not spend much time and effort to reproduce them. As a student, I once faced the issue of reproducibility, only to find out that, to be charitable, I was right and the computational paper I was comparing my computations to, had, shall we say, suboptimal results. In retrospect, the only error I made was not write a letter to the editor on the matter, it would have been a reference at a minimum. Then again, what's the point, we're talking about a field that really has not evolved for the past twenty years. If you want to read more on my thoughts about why reproducible research should be your main priority, check Nobody Cares About You and Your Algorithm.
In a totally different direction now, I have been asked this question by a few of you on how to cite Nuit Blanche as a reference.. I am not sure this is such wise thing but since the requests are there, let me give it a try by following these guidelines:
* For those blog posts that are entirely mine please use this for say an entry written on May 16, 2008:
Carron I. Nuit Blanche Blog [Internet]. Igor Carron, 2003 Nov - [cited 2008 May 16]. Available from: http://nuit-blanche.blogspot.com/.
* For those written by somebody else, please use the appropriate author(s) i.e.
Gemmeke J. Nuit Blanche Blog [Internet]. Igor Carron, 2003 Nov - [cited 2007 November 19]. Available from: http://nuit-blanche.blogspot.com/.
* For those of you who want to cite the Big Picture, here is my take:
Compressive Sensing: The Big Picture [Internet]. Igor Carron, 2009. Available from: http://sites.google.com/site/igorcarron2/cs
Jan 31, 2011, Postdoc, Technion We are seeking a highly motivated, talented, and fun post-doctorate fellow, for exciting research in the area of compressed sensing applied to analog signals with applications to low-rate analog-to-digital conversion, ultrasound imaging, radar systems, cognitive radio and more.Finally, two series of talks:
The research will involve tight interaction between engineering (signal processing and hardware
design) and applied mathematics and statistics. This project provides a unique opportunity to conduct research on topics that are in the frontier of signal processing and sampling theory, while at the same time have direct impact on industrial applications. Our labs are equipped with state-of-the art equipment from several companies interested in advanced sampling techniques for next generation radar/ultrasound/wireless systems and more. See http://webee.technion.ac.il/Sites/People/YoninaEldar/Info/hardware.html for a description of some current projects.
Potential candidates should have:
* An established research record with solid publications and the ability to work independently;
* Strong background in applied mathematics (relevant specialties are: sampling theory, optimization, signal processing, harmonic analysis);
* Knowledge of RF theory and design is a big advantage (but not necessary);
* A true desire to collaborate and be part of a team which will include both mathematicians and
analog design experts.
This project will offer the opportunity to collaborate closely with renowned experts in a stimulating intellectual environment, including the latest equipment in RF, radar, and other areas.
The working conditions at the Technion are excellent. The Department of Electrical Engineering of the Technion is ranked among the "top 10" Electrical Engineering and Computer Science departments in the world. The department is the major source of engineers who lead the development of advanced Israeli technology in the fields of electronics, computers and communications. A recent international review ranked the labs, the projects, and the student quality as "second to none". The Technion is located in Haifa, a beautiful vibrant port-city on the Mediterranean.
Details:
Starting time: Spring/Summer 2011 and onwards.
Duration: An initial contract of 1 year duration will be offered to the successful candidate, with a
possibility of extension.
Interested candidates are invited to send their CV with names of 2 references to Prof. Yonina Eldar at yonina@ee.technion.ac.il
http://webee.technion.ac.il/Sites/People/YoninaEldar/
Emmanuel Candés, Anders Hansen, Carola-Bibiane Schönlieb, Vincent Rivoirard, Jared Tanner will give talks on compressed sensing at Cambridge from March 21 till the 25th.
and (the day before yesterday)
Colorado State University’s Information Science and Technology Center (ISTeC) presents two lectures by Georgios Giannakis, Ph.D., ADC Endowed Chair Professor, Director, Digital Technology Center, Department of Electrical and Computer Engineering, University of Minnesota.
Both lectures are open to the public.
The first event is a ISTeC Distinguished Lecture in conjunction with the Electrical and Computer Engineering Department and Computer Science Department Seminar Series.
Wireless Cognitive Communications
* Monday, Jan. 31
* Reception: 10:30 a.m.
* Lecture: 11-noon
* Location: Lory Student Center, Room 205
Abstract
Exciting research and development efforts provide ample testament to the fact that wireless cognitive radio (CR) technology holds great promise to address fruitfully the perceived dilemma of bandwidth under-utilization versus spectrum scarcity, which has rendered the current fixed-access communication networks inefficient.
Accordingly, the need arises for intelligent radios equipped with critical cognition infrastructure to sense, learn, and adapt to their operational radio frequency (RF) ambiance. This talk outlines such an infrastructure for comprehensive situation awareness using the novel notion of RF cartography, which amounts to constructing maps capturing the distribution of power across space, time, and frequency; as well as the propagation medium per frequency from each node to any point in space and time.
Mimicking the way we rely on AAA or GPS-generated maps to navigate our cars, the vision is to utilize these maps for:
1. Identification of opportunistically available bands
2. Localization, and tracking of other radios
3. Interference control, resource allocation, and information routing
Credit: Atomization of a liquid jet, Institut de Recherche sur les Phénomènes Hors Equilibres
No comments:
Post a Comment