Erdős Number

According to wikipedia, "Erdős number describes the 'collaborative distance' between a person and Paul Erdős, as measured by authorship of mathematical papers." So, who is this Paul Erdős and why is his number famous?

Paul Erdős was a famous mathematician who had an enormous output of mathematical papers. Because of his enormous output, his friends created a metric as a tribute to him. According to this metric, Paul Erdős gets the Erdős number 0. Anyone who has co-authored a paper with him gets an Erdős Number 1. Anyone who has co-authored a paper with a person having an Erdős Number 1 and not directly with Paul Erdős, gets an Erdős Number 2 and so on.

Why am I telling you all this? Because, I believed that I had an Erdős Number of 4! I recently wrote a paper with my Professor, Dr. Ramakrishna Kakarala, who has an Erdős number of 3. This got me excited! The complete tree being: Vittal -> Kakarala -> Kelttee -> J. H. Straight -> Erdős. Therefore, my Erdős Number should be 4, right?

If you were a keen reader, you would have observed that I mentioned "I believed" and not "I believe". The reason being, I was wrong! I do not have an Erdős Number of 4 though I have collaborated with my professor who has an Erdős Number of 3. I missed a crucial part of the definition of the Erdős Number. "Erdős number describes the 'collaborative distance' between a person and Paul Erdős, as measured by authorship of mathematical papers." There you have it. Sadly, the paper that I wrote was not a mathematical paper. For a paper to be a mathematical paper, it has to be published in journals or conferences dedicated to mathematics.

Its funny how people can go from a state of euphoria to a state of despondency within a matter of seconds! Ah, but this is life...

CVPR 2010

This year's CVPR was nothing short of a major milestone in SCE's history. There were quite a few remarkable achievements in the top most conference in the field of Computer Vision. Here is the refreshing mail that greeted us today morning.

Hi all,

I have some wonderful news to share, fresh from here at CVPR’2010 in San Francisco:

SCE student Duan Lixin, with Xu Dong and Ivor Tsang, has just won the Best Student Paper Prize at CVPR’2010!!!

In addition, SCE has produced 3 oral papers at this CVPR, in terms of 1st author affiliation. This ranks SCE/NTU second among all organizations in the world as far as the number of oral papers is concerned. We are just behind ETH Zurich which has 4 oral papers, and we beat traditional CVPR powerhouses such as CMU, Stanford, MIT, Columbia, U Washington, Microsoft Research (all labs combined), Berkeley, Oxford, Cambridge, TU Graz, Technion, Weizmann and many others including NUS!

Furthermore, SCE has produced 6 poster papers for a total of 9 papers, which is a bumper crop compared to our numbers in the past (1 or 2, some 4 years ago and earlier)! I’ve heard various well known people remarking on the strong presence of NTU (mainly SCE, I think there is only 1 other poster from EEE) at this year’s CVPR.

The IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR) is one of the triumvirate top conferences in computer vision. It is very competitive with an overall acceptance rate this year of 26.7%, of which oral papers have an extremely small acceptance rate of 4.5%. Despite having only 461 papers at the main conference this year, it is heavily attended with some 1900 registered attendees.

Tat Jen

(with CVPR’10 published SCE faculty of Dong, Ivor, Deepu, Maylor, Clement, Ramakrishna, Jianmin and Jianfei)