Confessions of a researchaholic

December 27, 2009


Filed under: Real — liyiwei @ 6:00 am

Human life span is limited, but one could pass on his legacy to the future generations via two means: genes or memes.

The genes carry biological codes inherited to offspring, whereas the memes carry cultural practices transmitted through minds.
Both are fundamental mediums that shape human behavior. And more often than not, one would obtain most genes and memes from his family elders.

I owe my ancestors for these nice genes that they managed to survive and pass on to me, and my parents and grandparents for these good memes that they taught me by setting up good examples.

My grandfather was quite a unique figure, having been both a (Nationalist) air force officer and a classic Chinese scholar publishing articles and teaching in universities. This gave him a rare character combination: scholarly intelligence and military discipline. Through him, I not only learned how to have each but also how powerful such a combination could be.

I never felt emotionally sad for his pass away, not only because he had lived a heck of life, but also that I know he has passed on his legacy through both his excellent genes and memes. For me, the best way to memorize him is to continue his legacy.

(In memorial of my grandfather who passed away on Dec 27 2005 at age 88.)

December 15, 2009

Toon shading

Filed under: Imaginary,Real — liyiwei @ 6:52 pm
Tags: ,


Appleseed: Ex Machina is the best toon shaded animation I have ever seen. It really beats 2D cel animation.
I am too busy right now to write down more details, but just checkout this movie if you like anime or computer graphics.

Also, don’t miss the bonus features. I find it particularly interesting that the American crew talked about the “amazing collaboration” between Chinese (John Woo, producer) and Japanese (Shinji Aramaki, director) as though these two countries ought to start the third world war instead of collaborating on animation projects.

December 12, 2009

How to give a research presentation?

Filed under: Real — liyiwei @ 12:54 pm
Tags: ,

This is a vast subject, and probably has been covered by many articles (or even books). But here let me focus on the most crucial and fundamental issue.

The nature of a research presentation is to convey ideas that you know to others who do not yet know.

Corollary: no need for stage fear
There was a study indicating that people fear public speaking more than death. Not everyone has this issue, but if you happen to be nervous about an upcoming presentation, just remember the basic theorem: you know much more than your audience about what you plan to talk about. So, even if you make mistakes, the audience would probably not find out anyway. Just stay cool, and do not let your tone, facial expression or body language disclose the fact that you are screwing up.

Corollary: never over-estimate your audience
Unless you could read mind, it is probably very difficult to devine what other people do not know. Since you already know what you want to talk about, it is very tempting for you to recite what is already in your mind instead of what your audiences really need. This is the single most important cause for a bad presentation in my personal observation. There is no magic solution, but a useful heuristic is to never over-estimate your audience. Always start your presentation with the basics; if the audiences do not know that, they will appreciate your effort, and if they do, they will feel they are smart. Convey high level information instead of details, and use easily absorbable medium, like images or videos, instead of texts, to convey your points.

Corollary: never over sell
The goal of the presentation is more to entice people to be interested in your research than to teach them the details. In a sense, it is more like an advertisement than a class room teaching. It is nearly impossible for average humans to learn a new subject within a 20 min presentation. So do not try to cram in every single detail of your algorithm into your talk; probably nobody is going to get that anyway. Instead, focus on getting the audiences’ attention for the first 10 minutes of your talk. And if you could achieve that, you are on your way for a great presentation.

Finally, like many other aspects of research, the best way to learn is by experience. The more presentations you give, the more likely you will learn how to give a great one. Do not worry about failures; I totally blew up my first research presentation, but eventually I figured out the deals, and now I am not only highly comfortable but also highly enjoy giving research talk, especially to huge audiences like SIGGRAPH.

Theme: Rubric. Get a free blog at