Face tracking

I was really surprised when I saw this video of Logitech’s Orbit MP webcam. It’s about $100, and it has impressive face tracking algorithm. I knew that people were making headway in face tracking algorithms, but I didn’t think it was this good.

There are now open source tool kits for augmented reality. However, they use a template black and white pattern as a reference platform. It is definitely possible to use people’s faces as a tracking platform. Then, you can put statuses above people’s heads, like their names. Or in a augmented reality game, their “hit points”.

Incorrect predictions

One thing that really gets on my nerves is when people are dismissive.

“The wireless music box has no imaginable commercial value. Who would pay for a message sent to no one in particular?” — Associates of David Sarnoff responding to the latter’s call for investment in the radio in 1921.

Hindsight makes these predictions seems ridiculous, especially when taken out of context of the times and what was known as “common sense” or “what any educated person would know.” Unfortunately, people have only have their own experiences and expertise to draw upon. The model and framework by which we can evaluate and make sense of something new is by what we know.

“Heavier-than-air flying machines are impossible.” — Lord Kelvin, British mathematician and physicist, president of the British Royal Society, 1895.

“X-rays will prove to be a hoax.” — Lord Kelvin, President of the Royal Society, 1883.

“There is nothing new to be discovered in physics now; All that remains is more and more precise measurement.” — Lord Kelvin, speaking to the British Association for the Advancement of Science, 1900.

“Radio has no future.” — Lord Kelvin, Scottish mathematician and physicist, former president of the Royal Society, 1897.

Lord Kelvin was a major contributor to physics and engineering. I mean, he has a unit of temperature named after him. It’s just that he operated on the framework that he knew of. It’s very very hard to jump out of what you know and start all over. But some people just don’t event try.

Sketchup redos what’s old

It seems like much of the time, when something old is well done already, someone comes along and makes it new. Sketchup, at a preliminary glance is what one would use instead of CAD. It allows you to build and create 3D objects fairly easily and intuitively, as if you were manipulating it in space with your hands.

It recently got aquired by Google. So speculation is amok as to what this means for Google’s overall strategy. Personally, I think Google is big enough, like Microsoft, and Sun, to have their hands in everything. However, they seem to have a better track record than Sun so far when it comes to making their products work well as a thematic and cohesive whole.

The previous aquisitions that haven’t yet made a showing on Googles’ product lists are the things related to real-world computing. Things such as mobile socialization and traffic information haven’t yet surfaced. In addition, a friend of mine was approached by Google for a job in haptics (force feedback in robotic surgery). When looking at their job listings briefly, they do also invest in bioengineering and chemical fields.

What will Google use SketchUp for? No one knows. But I know this: that the easier people can create information–whether it’s written word, images, sounds, etc–the more that type of information is going to proliferate. If it’s really easy to post entries to a blog, many blogs will spring up. If it’s really easy to take photos, many photo sharing sites will spring up. If it’s really easy to make 3D objects with SketchUp, the more object manufacturing sites will spring up. Thus positioned, Google can continue to offer services which filter and organize that information; and that will contribute to this advertising revenue. But now, it seems to be in the position to also encourage people to create that information.

Phone form factor

I’d been saying for two years now that phones should have its buttons up top and the screen below it, since most people cradle their phones in their palms. In order to type on a phone like the RAZR, you’d need to balance it in with your fingertips.

Finally, Wildseed has the first phone I’ve seen with this form factor. Admittingly, it looks a little bit odd, since it looks like it was bent quizzingly. But it should fit better from a usability point of view. Form over function!

Matrix Photos in an Ikea Kitchen

IKEA | DREAM KITCHENS FOR EVERYONE is a flash site that uses Matrix-like action shots to introduce all angles of a kitchen to would-be buyers.

Not that I’m particularly interested in kitchens, but it does get me thinking about 3D photos. I think the way IKEA flash designers did it was that they had a studio with two or three dozen cameras set up in a semi-circle. And then they took an action shot all at once. Then, they use flash to stitch together the pictures so that it would look like you’re traversing in space, while time is frozen.

I remember my dad buying a 4 lens camera a long time ago. If I’m not mistaken, it’s a crude plenoptic camera. A real plenoptic camera has one lens, and has a microarray of lenses. Anyway, my dad’s camera took four photographs at once, so that when you go to develop it, you can get an actual sense of depth in the pictures. Or rather, it looked like people cut 4 layers of photos together, and put it in a hologram of sorts. The film was expensive to develop, and I think we only had a couple developed.

However, with digital technology, I think it is wholly possible with just one lens. Given that a lens can change focus fast enough, you can generate images that have different depth images. I haven’t read his paper, but that’s what Ren Ng’s Light Field Array camera seems to do. It can focus on different planes, AFTER you take the picture. There’s also Ramesh Raskar’s Line drawing camera. It uses four flashes and uses the shadows in the flashes to create line drawings and videos.

However, unlike matrix photos, there’s seems to be no information captured by a CCD array about the backside of an object when you take a picture of it.

And yet, I wonder. The reason why we can see anything at all is because light is reflected off that object. Sometimes, light bounces off multiple objects before it reaches our eyes. And yet, we only have information about the last object that the light bounced off of. I don’t know enough about optics (maybe it’s time to learn), but could it be possible that the light that reaches our eyes/camera still has information about all objects it has bounced off of? Unless the object is a mirror, light will scatter off of the object, so we will probably get partial information. However, if we were able to collect a bunch of partial information from different sources, it might be possible to piece together the partial information for a cohesive picture of the backside of an object, not unlike how network coding pieces packets together at the sink.

So if light still carried information about an object after the object has scattered it, and if we can detect it, I’m guessing it’s possible to take matrix photos from a single vantage point.

Internet Appliances

Interaction Ivrea: Gallery: Fluidtime

I’ve long thought that Laundry machines should be the first appliances to use the internet. I always forget that I have my clothes in the wash, and that I need to move them into the dryer. This was especially the case with living in the dorms. I wished my laundry machine would text message me when things were done. Now, I hear that there are people doing this (yay!), and that some college dorms have it.

What are some of the other things in the house that would be more useful with an internet connection? A toliet paper dispenser with an RSS feed and a toaster that can toast the weather on your bread are some that come to mind. If it moves in that direction, I imagine that either devices will get far too complicated, or that the design and use of everyday things will reign in the rampant featuritis that afficts those that make consumer electronics.

There are a couple things that have to happen before something like this is a reality. Home network is already cheap, but ethernet wiring in the house is not yet ubiquitous. Thanks to wireless routers, however, this will be a possibility, but will also raise the cost of these appliances.

Personal devices, such as cell phones, will have to have the ability to easily share meta information about its owner to these appliances. This will make it much more appealing for users to use internet appliances…it will be catered towards them.

Between perfection and performance

The psychology of learning:

“The people in the category perfection-oriented have a natural intellectual curiosity. They are constantly searching for better ways of doing things, new methods, new tools. They search for perfection, but they take pleasure in the search itself, knowing perfectly well that perfection can not be accomplished. To the people in this category, failure is a normal part of the strive for perfection. In fact, failure gives a deeper understanding of why a particular path was unsuccessful, making it possible to avoid similar paths in the future.

The people in the category performance-oriented on the contrary, do not at all strive for perfection. Instead they have a need to achieve performance immediately. Such performance leaves no time for intellectual curiosity. Instead, techniques already known to them must be applied to solve problems. To these people, failure is a disaster whose sole feature is to harm instant performance.” [emphasis mine]

I have to admit, though I’d like to think that I’m the former, I recognize that, in my work, I am mostly the latter. I like building things because I like to see them work, rather than the procedure that gets me there. So when things aren’t working, I’ll get frustrated debugging, and just start stabbing randomly in the dark. Then a couple hours later, I’ll figure it out, and proclaim that someone should “give me back my youth.”

I’ve been trying to change my programming habits so that it’s the former. I once lamented to IXM that I’m taking too much time getting going; to which he quipped, “Well, it’s the learning curve. As long as you’re learning something, it’s ok.” Rarely do people say something that make me feel better, but this one made sense.

However, there’s another side to this. Sometimes, I will be the perfectionist, and go to degrees that I shouldn’t, because the aspect of the work that I’m being a perfectionist at doesn’t really matter in the visible short or long term. It only bothered me aesthetically. In this case, I have to remind myself to move on, or else I’ll be shaving yaks soon.

I made a table.


So this past week, I built a table, so I can solder on it. I thought about buying a $80 workbench, which would have been really nice, but I thought I could:

  1. Learn a little bit about furniture making.
  2. Learn the value of tables.
  3. Have the oppurtunity to make something.
  4. Try to save some money.

A table should be simple to build, right? What’s the minimal traits of a table? This was was easy: A board with at least three legs. Of course, nothing is ever as simple as it looks, as the devil is always in the details. Since I was limited in my tool set–mostly hand tools, I had to make it as simple as possible. I went to Home Depot, and found out that they have pre-cut panels of wood. I chose a 2′ x 4′ x 0.75′ pine/particle board for $15, built legs out of 2″ x 2″ x 28″ of wood, and some top plates for $2.

The table is a little shaky, since I used my eye to judge the drilling of the hole for the metal dowels. But I think I can remedy that with some reinforcement. All in all, the raw materials only cost me under $30, and I’m pretty satisfied with it–as far as a soldering table goes. I had to spend more money to buy epoxy, a hacksaw, sandpaper, and two clamps, but I tend to think of it as and investment for future tool use. I already had a desk lamp, and a power strip, so you can see that I attached them to my table in the picture above. This is all I need for now, so there’s no use to making it until I really need it.

I think it’s important to take time to make your workspace comfortable and condusive to how you work. I tend to like everything all around me in a swivel chair, so I’ve built my office around that premise. I feel like too many people don’t spend enough time thinking about how their workspace should be so that it can induce good work efficiency. So now, my workspace is more comfortable, and I have another place to work. Eventually, I’ll have to get a real workbench with drawers and shelving. But for now, this will do. 🙂

Quantum-like effects on the Web

The article talks about how search engines affect the type of information that’s out.

Google, for example, says its mission is “to organize the world’s information and make it universally accessible and useful.” The way that’s written, one thinks perhaps of a satellite orbiting high above the earth, capturing all its information but interfering with nothing.

In fact, search engines are more like a TV camera crew let loose in the middle of a crowd of rowdy fans after a game. Seeing the camera, everyone acts boorishly and jostles to get in front. The act of observing something changes it.

Joel on Software had written a similar article four years ago about trying to manage software…by measuring it, you’ll change it. In both cases, the change comes about if there is feedback, or meta-data available to the subjects. If the subjects know what the observations are, and how the observations will affect them, they will change their behavior. It is much like quantum particles on the surface; their behavior changes when they get observed in the two slit experiment.

I seem to remember a similar effect in psychology experiements, but I don’t remember the name. In medicine, it’s the well-known, but little understood, placebo effect. However, they seemed to have gotten around it by making the experiments double-blind, effectively cutting off feedback to the subject.

This seems to work if the subject does not need to know the results of the observation. However, if the observations are direct tools in use by the subject, I’m not sure if one would be able to disentangle them from each other.

Though it’s by analogy, could the quantum weirdness be explained by a feedback mechanism? Could the underlying nature of physics be based on some type of meta-data that all particles have access to? I don’t know enough about physics to know where to begin looking, but it certainly makes me curious if any physicist has looked into it.