Project Recog

Imagine Cup calling..

2007-07-24T13:34:00.001+05:30

Its been a while since we've had a post here. We've been busy with work (of all sorts). Just about a week left for us to pack our bags to Korea and as usual, it is the last week during which all the crucial work is being done.

The people at Microsoft have also been busy with the organization of the event. You can check out the updated Imagine Cup website for information about the other teams who will be participating. (check out Team BlueLeaf from India). There are some really good projects and teams, and we're looking forward to be interacting with such a crowd (geeky crowd i must say) on an international stage.

Some good news we've got to convey is that there is another team from our college also coming to Korea. Team Dimensions will be participating in the Embedded Challenge, and have been shortlisted in the final 15. Way to go!

We're working on our software and the presentation, and hopefully it will all get wrapped up soon. The updated website content will be uploaded in a few days time, taking into account the impending competition. We also plan to release the alpha version here so that some of you out there can also tinker around with Recog to find out how easy it is to deal with the information explosion that you're subjected to everyday.

We have absolutely enjoyed being part of the evolution of Recog and can't wait to see how big a part it can play in the lives of many of the future users.

Neural Networks : a beginner's notes

2007-07-08T12:45:00.001+05:30

Its not so difficult. I'll stop short of saying its easy, but its not difficult. And thats good news. (calming smiley)

During the implementation of a certain module of Recog, using Neural Networks turned out to be part of the design plan. It was a big deal for me because it was my first foray into the field of AI, and I had no idea where to start. I had my design requirements at hand, but no clue whatsoever about how to implement it. In my quest for understanding how to implement Neural Networks, I did encounter some pitfalls and also some helping hands, and I thought it would be useful if I recorded some of my findings here:

- As I discovered the hard way, borrowing a large number of books relating to ANNs (Artificial Neural Networks) from the library and trying to digest the mathematical formulae does not help.

- Spending time to read authors' lame attempts at citing analogies from real life to help you understand does not help you much.

- Trying to solve your design issue by using ANNs without knowing why exactly are you using it does not help. (You'll curse yourself later for not understanding what values to set for certain parameters. This is experience speaking - however small it might be.)

- Try to build your understanding part-by-part, and not shoving down the whole sandwich down your throat. Try to find the simplest explanation of what ANNs are. (open the wikipedia page on ANNs and scroll down. ignore the initial junk) Found it? Now read it. Done? Now read it again. Because there is no point in proceeding without knowing WHAT are the advantages/disadvantages of using ANNs. Does it fit the picture of your requirements? If yes, good. If not, move on. Read about real-life applications and the different learning paradigms (supervised, unsupervised, reinforcement).
Remember, using ANNs is not going to be the cool part in your project/app. The accuracy of final results (and the speed in which it is generated) is what makes it cool.

- Next, run to the Search Gods. If you think that there's going to be no solution implemented yet for the design problem that you're dealing with, think again. Chances are that you're very wrong. People (tons of them) all over the world are solving tons^tons of problems using ANNs, and almost all of them write a research paper on their solution (whether it works or not).
So, spend a lot of time on Microsoft Live Search Academic and Google Scholar to see if any kind of solution to your problem already exists
(I like the interface of the former more because it saves you trips to unnecessary pages by providing abstracts side-by-side).
Go through these papers (atleast half a dozen of them) and find out where certain approaches failed and where some succeeded.

- Finally, and most importantly, what I've learnt is that you've got to implement it to understand it. There are some Neural Network libraries available online which allow you to code and bring your neural network into actual existence. I'd suggest AForge and Brainnet libraries, both are C# .NET based libraries. Try downloading both of them, and see if any of the sample programs accompanying them resemble the method of solution to your design problem. The code is commented nicely and you won't have much problem in grasping the technique after a while.

I'd like to repeat this: you've got to implement it to understand it.

Hmm.. just remembered something. As I write this post, somewhere in Berkeley, people are trying to seek the tech beyond Neural Networks.
Something they claim is closer to mimicking the human brain.
Is that even possible? I believe so. But not yet... Not yet.

Also, just as I am about to finish writing this post, Vasu and Avi are trying to feed inputs to the neural network that we created (tested with few inputs yesterday and it worked!!! Now training it for a huge sample train. Fingers-crossed. Praying.)

Have fun connecting them neurons. They good.

The coming of Recog - Episode 1

2007-07-04T21:34:00.000+05:30

Argh! I am last Again! dj and vzu have already posted their blogs and I am sitting here and reading them :s. Now Its my turn...

We have not talked much about Recog on this blog yet. Although you'll get bits of information from elsewhere now and then, we would be revealing this a bit slowly :) we like the suspense part!...

Searching. Searching is something we all do for several minutes (and hours if you are like me) everyday. But are we all satisfied? Do we get the results for what we are searching the first time we search for it? or do we still have to type two or three different combinations of "keywords" to get the right result? Strangely, We notice that even Google does not live up to the expectation of giving you what you want.

Picture this:

You are doing a research on Science Fiction novels and looking at writing styles of various authors. You go to wikipedia coz that is the first link on Google. There you find references to H. G Wells, Issac Asimov and many other great writers. You also find that there is a different genre of "Mystery Fiction" and read about Michael Crichton.

Great!

Now you have about 15 tabs open and you have no place where you can read through all this content and at the same time arrange it and organize it for the 10 page long essay that you need to write for your "Approaches to Science Fiction" course. Any idea how you are going to manage all this? Yes that is where we come in.

With Recog, you write the introduction paragraph for your essay as you read the wiki page and other source material. Soon you write a bit about Asimov, Wells and Mary Shelly! By this time you have written a few paragraphs and Recog has done its job meanwhile! While you were writing the introduction page, it searched the net for various genres of Science Fiction and linked them to your text automatically. When you were writing about Crichton, It searched amazon for his latest books and brought to you the links for the books like State of Fear from the author, along with relevant book reviews. Recog also added the latest media related to the topics to your workspace! Having done all this you prepare your report, rearrange the content with simple drag and drop to arrange it to your liking. Voila! your work is done! And all this stays with you for future reference in case you want to look at it again!

So that was the searching in our Recog... where the software conducts intelligent searches giving you accurate and relevant results.

Remember this is just a small part of what we call Recog! The more important things come after some wait :)

Meanwhile, feel like talking to us? Reach us at:
< ping [at] projectrecog [dot] com >

My First one..

2007-07-04T19:33:00.000+05:30

Yeah, so here goes nothing.
I have been told so(ooo) many times to scribble my chatter on this blog (coz, yes, i do chatter a lot..)
And my persistant (lame) excuse used to be "I'll blog when I'v finished doing something (workwise)". So while the latter never happened, i made sure the former didnt as well. So you see..

Ok, lemme give u a peek into how this works. Dj is the pusher kinds - positive, 'we'll-do-it' and 'we'll-do-it-well' kinds. (He also comes up with lame jokes when Avi and I totally-totally take his case.) Avi is the quiet, conscientious nice guy kinds, who is ALSO good at taking cases of people. And he codes (and debugs) really well [understatement alert!].. (and therefore i bother him with my shoddy code ALL THE TIME! :D ) And I am the worrier of the team. I hop around and I worry. Yes, you read it right. Thats my job. (Ok, even though i hate to admit it, Avi and Dj, SOMEtimes and only sometimes, team up and are able to take my case.) And i think we make a fairly complete and well-balanced team. We function fine and its fun working - well, most of the time.

OK, this turned out to be a session on who is what-types and who-takes-whose-case, even though i didnt plan so.
More later.. (I like this, so i think i'll keep this coming)

Avi's sparetime work

2007-07-03T23:23:00.000+05:30

Here is a picture of what Avi is capable of when he has his creative bursts of energy. Not bad at all eh?

(thats his keyboard.. his life revolves around that. i wonder if its still working.)

In 30 days, Dynamic Korea...

2007-07-03T23:17:00.001+05:30

Thats right. Just about a month left for us to take Recog to the world stage. The Imagine Cup World Finals will be held in Korea between Aug 5-10, and there's a lot to complete before that. We have started with our visa procedures for entry into the kingdom of chaebols.
Recog is evolving day-by-day, and it should be all set by the end of this month. We also have our B.Tech project presentation coming up in a couple of weeks, and some research documentation has to be done for that.

Microsoft was supposed to release the information regarding all the other finalists for Software Design Invitational on July 1st, but no signs of that yet. Once that is out, we can add more info on this site about the project (at this point of time, its all classified )

Meanwhile, we're toying with the idea of giving a sneak preview of Recog to you all out there before the end of this month. Will keep you informed about that here. Keep visiting, and as always, your comments and suggestions are welcome.

About Us

2007-06-29T00:31:00.001+05:30

Hello All,

This is to let you all know that after a lot of waiting, we have updated the About Us page on our web-site. You can now find info about our team: who we are, where we are from, and what we do. Also added a couple of photographs + snippets of info about our Imagine Cup journey. More later..

Cheers!

Space!!

2007-06-19T15:04:00.001+05:30

And finally, we get a lab to work from. After weeks of running around from pillar-to-pillar of the college administration and faculty, we now have a room where the three of us can work together without any time restrictions. There were some faculty who wanted to control our working hours (can you believe that?!?), but it was a no-deal for us. We are nocturnal creatures, and it will remain that way to ensure that this project is completed in time.

The place that we are settled in right now is called HashLab.
Thanks to Prof. Binita Desai for helping us out in this regard.

Building the website

2007-06-19T14:44:00.000+05:30

Our main website is under construction as of now. So until fresh, relevant, interesting content is generated and processed, we shall be alive on the domain through this blog.
We're also working on some other pages which will be updated in a couple of days.
(to be more specific, we had assigned the 'About Us' page to Avi first, who took some time and decided that it should be done by someone else. Then, Vasudhara tried working on it, and she said she tried 'different' approaches, none seemed to work out)
And that leaves me.. to save the page. I too havent exactly been punctual about it, already running late by a couple of days. But it'll be up soon. I promise :)

Welcome to Blog Recog

2007-05-28T02:17:00.000+05:30

This is the place where we update our audience with all the news regarding our project Recog. There is still a long way to go in the development of the solution, but we believe it can make a difference to the lives of many people. Will keep you all informed of the proceedings, and looking forward to your valuable comments.
Team BlueLeaf is at work. :)