Computer Vision and supernatural species by Peter Tu

This interesting article is from the blog of Tomasz Malisiewicz. It is a blog of Peter Tu (see the original article ).

 

Peter Tu is from Computer VisionGroup of ge Global Research.

Tu's analogy between researchers from academia and industry with those from the werewolf, vampire, and undead supernatural species is true.

In fact, some problems were also explained in the joke. Therefore, Liao was inspired and entitled to be entertaining.

Below is the original text + simple excerpt.

A few weeks ago I attended two conferences: the European Conference on Computer Vision (ECCV) and the IEEE Conference onAdvanced Video and Signal-Based Surveillance (AVSS ). at both events I wasinvited to particle on the vision and industry discussion panels, where weconsidered how computer vision research relates to the business community. during these debates it became clear to me that academic researchers or theProfessors and Industrial Researchers (IRs) view themselves as distinctlydifferent breeds. I came to the conclusion that there is a good analogy betweenComputer Vision researchers and the supernatural.

 

Peter Tu came up with a point of view after discussing the visual industry group in ECCV and AVSS: academic research and industrial research can be analogous to two different supernatural species (vampire and werewolf ).

 

Let's start with the Professors, whoseanalog wocould be vampires (vampire ). vampirestend to be solitary creatures (solitary species) skulkingaround in their musty castles. professors also tend to keep to themselves, but insteadof castles they are found lurking in universities. vampires need blood fromhuman donors. professors need publications, which they extract from their gradstudents. while vampires are immortal, Professors have tenure.

 

Professors is like a zombie, but the vampire is hidden in the castle, and Professors is in various Universities. Just as a vampire needs to suck people's blood, Professors needs to extract the labor force of graduate students to obtain publications. The vampire is not dead, but the professor has a term of office.

 

Vampire and academic societies are bothvery hierarchical. the position of a given actuator sor can be directly inferredby observing where he or she sits during the conference dinners. manyenthralled mortals (confusing mortal) hope tobecome vampires, but few ever do. likewise implements grad students hope to becomeProfessors, however at any given time one can count on one's fingers the numberof open computer vision Professorships. export grad students end up drifting awayfrom the field all together. some end up on Wall Street solving differentialequations for their between ER and money-lending masters. however others canmake the conversion to industry, where they become the research equivalent ofwerewolves.

 

There is another similarity between the vampire group and the academic community, which is classified into different levels. Just like many confused people who want to become a vampire, many graduate students want to become a MongoDB, but only a few of them can. Some of them are engaged in the financial industry, while others are in the industry and become a "Wolf ".

 

Werewolves have larger physiques whencompared to the diminutive Vampires. in the research world muscle is equated togear and big working systems, which IRs spend much of their time collecting andbuilding. werewolves live in packs, for IRs such communities are known asw.ate research labs.

 

In the industry, the "werewolf" needs strong muscles and physique, which are the production chain systems that industry researchers spend a lot of time building. The "werewolf" is a group, and their group space is the research labs of major companies.

 

Like Vampires, Professors tend to havearistocratic tastes and interests such as how to make proper measures onRiemannian manifolds, in contrast werewolves and IRs are a little earthier-IRs will talk at length about such issues as runtime and robustness.

 

Professors have a taste similar to that of the vampire's nobility. For example, they consider how to make a strict measurement in the popular column, while the industry's "werewolf" concerns are relatively vulgar, they are concerned about the operational efficiency and robustness.

 

Werewolves must periodically make ametamorphosis into hideous monsters. IRs are equally cursed. instead of thefull moon they have quarterly reviews. during such distressing times they mustput down their slide rules and compasses to become power-point spewing1_ate stooges. for senior IRs the transformation can be so complete thatthey can be seen dawning both jacket and tie!

 

The werewolf becomes a violent beast every night of the moon. Similarly, industry researchers also have their "Big Aunt", which is only quarterly. Every time a quarterly report is made, they have to put down the "calculation ruler" and "circular gauge", instead of focusing on the production of Power-point.

 

There has always been animosity betweenvampires and werewolves. professors and IRS also share a certain amount ofcontempt for each other. professors feel that they are the only one's fit toconsider fundamental problems and that once they have figured things out, theproblems can then be sent to IRS who shocould be relegated to working out thefine details. while IRS assert that the professors simply play around und withproblems, come up with solutions that don't really work, subsequently claim theproblem to be no longer interesting and then move on the next topic de jour. like all prejudices, both views are neither charitable nor accurate.

 

The "werewolf" and "vampire" are hostile to each other. Professors and industry researchers also interact with each other. Professors think that they can solve the basic problem, and then they will only fill in details when getting the industry. Researchers in the industry believe that the professors only give a try to solve the problem, and the obtained solutions do not work in practice. Of course, they are all one-sided.

 

Wherever there are vampires and werewolvesthere too are vampire and werewolf hunters. researchers also have their nemesesand they are known as Masters of Business Administration (MBAs ). instead ofstakes and silver bullets, MBAs come armed with spreadsheets, schedules andmarket surveys. their zealous attacks can come from any angle includingquestions regarding relevance to society, projections of customer return oninvestment and concerns regarding the timeliness of results.

 

In the movies, the young starlets alwaysfall in love with the supernatural heartthrob-well I guess all analogies havetheir breaking points... As I mentioned at the beginning of this piece I recentlyattended a convening of computer vision researchers. At all such events one mustflex a little muscle-see the video below...

 

...... It's time to get back to the topic. As I started, I attended two CV meetings and showed you something amazing, please see the Demo Video below ......

 

(The video is on YouTube. If you are interested, you can view it on the wall)

 

It mainly involves:

Action recognition, action analysis, behavioranalysis, behavior recognition, tracking, facial analysis, biometrics

I am very familiar with it. I believe many laboratories in China are also doing related things.

Several images: