Tech-watches as the new mobile device?

You can call me a geek, but I definitely had a watch like this when I was a kid (image from wikipedia):

I actually had a better one, with a remote control device. It could learn the IR signal from a hand-held remote control and send commands to control any device. It was a miracle and I loved it. I remember annoying one of our teachers at school by messing with the TV/video setup during an AV presentation.

The NYtimes has an interesting article about how tech-watches have simply missed the boat in terms of personal devices. The article mentions some companies that are looking at creating watches that are either replacements or companions to smartphones.

Do they have cameras? Not yet, but they could. They would need to be really low-power. There are also a lot of privacy issues with a person carrying a (possibly) live, uncovered camera on them at all times.

The article sounds skeptical, and I can understand why. Some of the descriptions of possible niches for tech watches seem pretty contrived. The worst is a watch that syncs with your phone, and basically filters certain content that you might want to keep track of. That would really be a meta device, that perhaps constantly keeps track of high-frequency content that is too trivial for you to keep taking your phone out (email, twitter perhaps?).

Also I like the security aspect of it. I find the prospect of losing a smartphone through carelessness to be frightening. If my phone was safely strapped in, while my watch allowed me to limit the number of times I took it out, that would be good.

Micro Medical Vision

The Economist has an interesting article about efforts to miniaturize medical imaging systems. Scientists have managed to make a PET scan for rats. The animals can actually wear the devices, and the images are broadcast over periods of time.

First a few comments about medical imaging and medical computer vision. Medical vision is a vibrant field. I'd like to count it within computer vision, but I'm sure people from medical vision might contest that: there are separate conferences and workshops that are quite famous. We in traditional computer vision may be ignorant of the latest in medical imaging, but we know and respect the use of heavy mathematical machinery (applied physics, machine learning and topology) as well as the maturity of technology (Siemens, GE and others build impressive, working medical imaging systems).

Although I'm not up to speed with the latest in the field, I still feel that everyone should be amazed that PET scanners could be built so small. The only similar thing I know of (from my time at CMU) is George Stetten's hand-held ultrasound , which was a major breakthrough.

The mind boggles with the possibilities for medicine in remote locations and the analysis of miniature biological entities like insects (modulo radiation concerns).

Multitalented Amazon

So when I first came to the United States, in 1999, the CEO of Amazon, Jeff Bezos, won Time Person of the Year award. It was a heady time, but as we all know, the dot com bubble exploded.

Amazon survived, and built a safe rep as an online superstore.

Then something happened. Perhaps it was always a plan, but Amazon started exploring new horizons, and putting out new products and services that didn't fit with the concept of a boring online bookstore.

I'll quickly mention two of these. The first is mechanical turk (article by nytimes), which is the jewel of the crowdsourcing community. Here Amazon is leveraging its online cloud infrastructure to create a market place for crowdsourcing. The second is the Kindle, where Amazon is betting on low-end personal e-readers. Here Amazon is leveraging its position as the premier online bookstore to provide the best titles.

First, I think this is interesting because it shows Amazon has ambition and wants to be a tech company that produces hardware and software and is not stuck in a niche. Secondly, its only a matter of time before Kindles get cameras, whose needs might be low power consumption and reliability. In addition, computer vision people also are using Turk for a variety of applications.

Interesting times ahead for those of us who track Amazon.

Curved electronics

A lot of my recent research has been on creating optics that can "see" over a wide field-of-view. The main issue is that for conventional electronics, projecting a wide FOV onto a planar imaging sensor creates perspective distortions.

This disadvantage is only one of many that exist with traditional circuitry, since we are restricted to electronics that are created on flat silicon sheets.

Curved, stretchable electronics are coming though.

Prof John Rogers of UIUC and Prof George Whitesides of Harvard have started a company to build curved electronics. If you have seen their papers, then you'll know they have already got a number of interesting applications, such as a curved image sensor to remove spherical distortions.

My interest is in how these will enable flexible, miniature sensors. For now, however, the size of each sensing element ("pixel") is much larger than what we can get with flat CCDs. Hopefully these will reduce over time.

Optical templates made of silk?

The Nytimes has a really interesting article about silk. Its mostly a material science article, with tons of interesting applications.

However, one of the possible directions the silk scientists could go in, is putting millions of tiny mirrors and/or holes onto clear silk. This would allow manipulation of light at very fine granularity.

HP labs releases annual report that discusses immersive 3D

HP labs has had a good 2010. From its annual report, it appears that over 120 "technologies" (possibly covering software, hardware, patents, processes etc?) have been transferred over to HP businesses.

While this is the propaganda from HP labs itself, and should be taken with a grain of salt, the reputation of the labs in the computer vision community at least is solid. More relevant to this blog, the report has a small section praising the possibilities of mobile 3D and explaining HP's take on the scene.

From what I understand, HP's approach is tightly coupled between displays and autostereoscopy. HP's display hardware research demos don't have anything supersurprising, since they include the usual "paper-like" thin display technologies with a focus on energy efficiency.

A tightly integrated approach between displays and 3D means that HP is thinking deeply about how mobile/immersive displays and 3D displays come together. The implication is that you might see HP products with paper-like, super-thin, autostereoscopic displays sometime soon. Mobile 3D might go massively immersive.