Showing posts with label IEEE. Show all posts
Showing posts with label IEEE. Show all posts

Sunday, August 30, 2009

Another school year begins

Ah, the wafting aromatic waves of beer and cardboard boxes could only mean one thing: the kids are back in town. Here are a few minor updates on my end of things (major for me, minor for readers, I suppose)...

Mendeley sadness
My schedule has become too packed and unpredictable to keep up with the Mendeley work, so I decided to leave my Community Liaison job (on good terms). I kept thinking, "As soon as I get over this next time crunch, I'll be ready to really give Mendeley the attention it deserves." But, those crunches just kept coming and I don't see them becoming any less crunchy in the near term. The folks there are great, the product is great, and I really wish I had the time to work on it more. I expect really big things out of that group, and will continue to sing their praises. Okay, maybe not sing. Limericks? Nah. Too many syllables in all the keywords ("There once was a bibliographic... shit, ran out of syllables already").

My Project/Thesis/Thing
I've begun running some of the experiments that I've taken lead on (I don't want to say they're 'my' projects because so much feedback has gone into the process that, like all good experiments, I don't think one person deserves all the credit). I would go into more detail, but you know how it is in research and especially FDA/clinical trial situations. Still, these are 'mah babies' and so far things are looking really good. Now to get to the data! Big thanks to Jessica (my 'med school minion') for putting up with my odd schedule and all the great work. Hrm, I could use another DNI contributor or two...

On the topic of putting together kick-ass experiments, I've been meaning to post that there is a whole field out there that many scientists are completely unaware of called "User Experience", or UX. UX is kinda an amalgamation of user interface, HCI, psychology, feedback, and design that looks at how all those fields converge into the subjective experience of the user. Fascinating stuff, though more on the 'web design' end of the spectrum, and I've been really loving the bits and pieces I've been coming across. This field is set to explode with the continually expanding emphasis being put on web apps, mobile apps, and even OSes in general. Which brings me to another nice tidbit of knowledge.

2 copies of Windows 7 for $30 (for students)

Yeah, right! No, really! Thinking of getting a new PC, but wanna wait for Windows 7? Don't want to reinstall your system halfway through the semester just play with all the fun new built in toys? (Shuddup Mac people. some of us prefer a different flavor of Kool Aid.) Well, you'll love this.

Join IEEE as a student for $30 (1-year membership), and get access to their MSDNAA software library for free. This includes the Windows 7 RTM version. RTM? That sounds like a beta. Nope, RTM stands for 'Release to Manufacturer'. Buy a PC on Oct 23, and this is same version of Windows 7 that will be shipping on it. Not beta, not RC. I've gone through this process, and it works as advertised. Buy the membership, wait about 7 days for the MSDNAA specific email to show up with your login and password (mine took about half that long, maybe because I used my university email address), and go get em, tiger.

Only the Business versions (32 and 64-bit) is up as of this posting (don't waste your time with the RC, and Ultimate isn't much better than Business, unless you need the language packs, which I think can be purchased separately). Supposedly, the 'in place' upgrading has undergone some major upgrades, and Vista users may not even need to do a clean install, though I always recommend it. Mine installed fine - took about 30 minutes and activated without issue.

Here are the linkies:
Source of the info: SlickDeals
IEEE description: here

Data Visualization
I plan to add more on this in the future, but an interesting (free) iPhone app to check out is Roambi, which basically allows you to upload 'data' to their site, and then spit is back in some really visually pleasing, interactive displays on the iPhone.
Try the app's built in demos for a better idea.

Tuesday, June 3, 2008

Hallelujah transhumanism!


Alright, I like pontificating on the future as much as other people, but really, tranhumanist types, the singularity stuff is getting too much attention and damaging legitimate attempts at more 'humble' science. I got thinking about this after reading the title of this Neurophilosophy post ("The Technological Rapture"). Isn't that what it is? Just another substitute for religion, but with nanobots and bytes instead of angels and miracles?

Yeah, it is, just with a more physical/material slant. Technology/God will come and sweep away the chosen ones that embrace it/him! Into a land of pure energy/spirit and knowledge/love! Those that choose to ignore progress/the lord will be left behind as the Earth is destroyed by man made disaster/demons or catastrophic solar event/rivers of blood and whatever!

What separates the two is the semi-feasibility of approaching a 'singularity' point where knowledge progresses faster than we can perceive it. Problem is... that happens NOW. 20% of the US has never sent an email. You have to pull your head out of math and science, and submerge it into sociology and philosophy a little here, kids. Or, if you need an equation/variable as a mental speed bump, how about this: Progress requires that the perceiver/creator both acquire knowledge and act on it (whether thinking/building/etc) in order to influence other perceivers/creators. Evolution dictates that organisms will act to increase fitness, therefore use of the aforementioned knowledge will be used to aid the individual with it. If speed of advancement is infinite, then the population instantaneously is cleaved to a single person/creator/etc with infinite lifespan (that kookie 'god moment' type thing). Any semi-intelligent person, say Mr #2 to Mr #1 above, would realize this and destroy Mr#1 in order to stay alive, as Mr#3 would gank Mr#2. The whole process converges to a "Mediocrilarity" where advancement cannot be faster than acting on it because the 'progressors' are held at some type of equilibrium point others can comprehend. Tada! Stalemate.

So, stop talking about brain implants like they'll be used to control planet-seized sentient robots that can time travel and fight angels with dark matter. It makes getting funding harder when everyone with a normal 9-5 job thinks you're wacko.

(Here are those recent articles on this. (shame on IEEE Spectrum for hosting this.))

Tuesday, May 6, 2008

The missing link


Behold, the memristor. Why is this not bigger news? Being touted as the fourth entry to holy trinity of electronics today (resistor, capacitor, inductor), the memristor retains information about its state without power.

Now stop. Seriously think about that for about 30 seconds. This means no boot up times, no spinning platters, zero power consumption sleep mode, no phantom power sucking consumer electronics, essentially on demand activity - you flip the switch and you're instantly at the previous state. Just like a light bulb. I would not be surprised to see a new power box and wall plugs for homes, where a central power distribution unit that is sent an 'on' signal over the line, that kicks electricity only to the devices that need it. More safe (juice not flowing all over the house to cause fires or zap unsuspecting lil ones), more efficient (all wire has some resistance), and it jsut plain makes sense.

Right now the EE software doesn't take advantage of this basic, elemental part, but I am excited to see what incredible devices will be built in the future. I wonder how long until the process is commercialized and available everywhere. With this discovery, HP essentially vaulted themselves into the mindspace once only occupied by IBM. (HP's had some great research, but this kicked it up a level.)

Oh, and for BCIs, the biggest problem for future real-world applications if we figured out all the crazy decoding stuff, is power. Tada! Solved.

Then again, our (USA) energy problem = solved, for now. Even if it costs more to produce, which are you going to buy: The huge honkin HDTV and PC that suck down $150/year when off, or the memristor based versions that cost $200 more?

Sorry, this is one of those news items that totally gets me juiced. Heh, juiced. I'm funny to me when I regress.

Wednesday, May 23, 2007

Catching up on old-ish news

Over the next day I'll be catching up on stories that slipped by me as I was setting this blog up. For tonight here are a few...
Our lab/CKI were mentioned in the media lately for our multiple papers are the recent IEEE conference in Hawaii. When our 'troops' returned for our usually Friday meeting, we asked how the meeting went. They started into how the weather was great and they went to the beach and played golf, but didn't realize we were asking how their presentations went. Ah, I tried to offer myself up to "fall on this grenade" but, alas, no data = no hula girls. From RedOrbit:


Three presentations on Saturday, May 5, 2007, by engineers from the Company and collaborators at Brown University highlighted advances in Cyberkinetics' BrainGate Neural Interface System included:


"Multi-State Decoding of Point-and-Click Control Signals from Motor Cortical Activity in a Human with tetraplegia." Authors include Sung-Phil Kim; John Simeral; Leigh Hochberg (VA/Harvard Medical School); John Donoghue; Gerhard Friehs; and Michael J. Black. All authors are affiliated with Brown University. This presentation described the development of "point and click" computer cursor control, which would be necessary to control medical and other devices, including wheelchairs; "Decoding Grasp Aperture from Motor-Cortical Population Activity." Authors include Panagiotis Artemiadis of the National Technical University of Athens; as well as Gregory Shakhnarovich; Carlos Vargas-Irwin; John Donoghue; and Michael J. Black, all of Brown University. This presentation described decoding neural activity to achieve "continuous motion", which would be important in the ability to provide more natural control of a person's own arm and hand movement, as well as prosthetic limbs; and "Automatic Spike Sorting for Real-Time Applications." Authors are Daniel Sebald, a consultant to Cyberkinetics, and Almut Branner, of Cyberkinetics. This presentation described automatic sorting of brain signals, which significantly reduces the time between a person's thought about movement and the ability to put the thought into action, which is aimed at providing "real-time" performance.

I am posting at 3am right now, which means I could really use a device that would just put me to sleep with the press of the button. Hrm, where could I find one of those? Oh, here!



Interestingly, is basically forces subcortical structures to into a delta oscillation using TMS and you just fall asleep. There is a series of rhythm changes they refer to as Sleep on Command, and a beta oscillation mode to aid student in studying. My lord. I could be the Ultimate Grad Student! Imagine a grad student that never slept! PIs around the world begin drooling. From PR.com...

The sequenced frequency system called Sleep on Command™ was first released to the public in May 2005 and provides clients with an extraordinary sleep experience. The Sleep on Command™ trademark was filed in early 2006. An Alert Mode (low Beta wave) program was added in 2006 to aid students in studying, and to aid drivers in maintaining alertness and reducing physical and mental fatigue on long trips. It works equally well in desk job
environments.

The device costs $500 and can be purchased directly from the company website (with a 1-year warranty). I think I might order one. *counting pennies* Actually, this is the gimmicky device version which you put under your mattress - so it doesn't penetrate your skull. And here I thought I had found a sleep on/off switch. *sigh*