Squobble Blog

Comparing the weather forecast for two cities

2011-11-15T12:29:00.005+01:00

Going out for the weekend but can't make up your mind between two great places? Wolfram Alpha to the rescue! I found out you can compare weather forecasts with it too. Just use their weather forecast with the "versus" operator, like in this recent query of mine

and up pop the results:

For less well-known cities you might give Wolfram Alpha a hand by adding some parentheses etc.:

Happy walphing!

An exquisit blank HTML page

2011-10-04T00:17:00.001+02:00

Fancy hacking together a quick single-page web site? No? Well I do. Just in case you change your mind, here's a high-quality blank HTML page you can use as a starting point. It's transitional XHTML and has utf-8 encoding, so that should allow for doing quite a few crazy things.

Spanish DNI and NIE letter calculator

2011-09-29T14:03:00.000+02:00

As I couldn't find this really well done anywhere online, I made a small tool to calculate the letter of the Spanish DNI. For expats living in Spain, you can use it to calculate the letter of a NIE too.

Here it is: Calculadora de letra de DNI y de NIE.

Google reverse image search on desktop

2011-06-17T10:27:00.002+02:00

Google finally adjusted their Goggles service for the desktop, so we can now also add their service to list of reverse image search tools, which already included the algorithms of TinEye, Cydral and GazoPa among others. First results show it's a winner, returning more identical results, more visually similar images, and interesting context information. Go to images.google.com to see if it's available in your region.

To commemorate the occasion, I released a new version of RevEye, the Chrome plugin to perform reverse image searches. It now includes searches by Google, TinEye, Cydral, and I've added Yandex and Baidu to the list since they also do pretty well.

RevEye v1.3 is a free download available from the chrome web store.

RevEye: Even more reverse image search

2011-06-01T12:15:00.007+02:00

Google asked me to change the name of my Chrome plugin "ChromeEye" into something that did not violate their terms. Fair enough. I changed it to "RevEye", and took the opportunity to make some changes to the user interface as well.

First, I added the Cydral engine to the list of reverse image search services included, since it has been producing some very nice results lately.

Second, I added an option to change the context menu: You can now choose either to use one default search engine, or to have a cascaded menu with all options on every search. My favorite option is the one that opens all services in different tabs.

These changes basically make RevEye a "meta reverse image search tool", since it now includes the complete functionalities by the TinEye, GaZoPa or Cydral plugins. So off you go then! Install it and have fun tracking images.

RevEye v1.2 is a free download available from the chrome web store.

More reverse image search: ChromeEye v1.0

2011-03-02T11:39:00.001+01:00

Yesterday I released a new version of my reverse image search extension "ChromeEye" for Google Chrome. While the first version was simply a bridge to the reverse image search engine TinEye, the new version 1.0 includes the option to use different search engines. At the moment that means images can be sent either to TinEye or to GazoPa. TinEye, which seems to be most popular reverse image search engine these days, looks mainly for exact matches. GazoPa uses a very different algorithm, based on visual features, with sometimes surprising results. Both are still improving their algorithms, obviously, so we´ll keep an eye on them.

That said, ChromeEye is now a plugin for reverse image search in general, and I will be including more search engines in the future. Note also that it is using the right click context menu for image searches. This was a much asked for option in the previous version, but technically not possible (due to Chrome restrictions) when I released it.

ChromeEye v1.0 is a free download available from the chrome extensions directory and from the chrome web store.

Fixed-Budget Kernel RLS at ICASSP 2010

2010-03-17T19:07:00.002+01:00

My paper titled "Fixed-Budget Kernel Recursive Least-Squares" will be presented this afternoon at the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2010). It was written in collaboration with Weifeng Liu and Jose C. Principe of the Computational NeuroEngineering Laboratory (CNEL) at the University of Florida. I couldn't be there to present the paper myself, but luckily my former supervisor Ignacio Santamaría is attending and will be giving the talk. If you are interested in learning methods or online kernel methods, be sure to attend. It's the first paper of the "Learning Theory and Models I" session (MLSP-L2.1), which is held on Wednesday, March 17, 13:30 - 15:30.

The basic idea behind this learning algorithm is pretty simple. First of all, it is an RLS method in kernel feature space, which means it is capable of learning a nonlinear regression in an online manner. Kernel RLS methods usually face memory and computational problems, since their solutions require to invert a large kernel matrix that incorporates information about all processed data. As new data patterns (x,y) arrive (remember this is set in an online environment), this matrix grows and the computation of the regression output becomes more costly in each step. The typical approach to avoid these memory and computational problems is to restrict the data patterns to a minimum number that allows to approximate the solution sufficiently well. Each time a newly arrived pattern is found to be sufficiently interesting, it is added to a "dictionary".

In contrast to these techniques, which make the dictionary expand (albeit slowly), the technique of this paper fixes the size of the dictionary. As a consequence, apart from adding patterns to the dictionary, we also have to prune patterns. While the pruning itself involves only simple algebra, the criterion to determine what data to prune is more interesting. We chose to use a relevance-based criterion in this paper, motivated by the good results it yielded. More specifically, in each time step we calculate how relevant each stored pattern is with respect to the regression, and subsequently prune the least relevant pattern. (Details and references are in the paper itself.) As a result, after running a number of iterations of this algorithm, it obtains a compact set of patterns that summarizes the observed patterns pretty well.

I particularly like the interpretation of this algorithm as a simple learning method that deals with memory restrictions. Every time this method receives a new data pattern, it must determine actively what pattern to "forget" in order to maintain its memory size. So it's a way of "optimal forgetting". And as the results in the paper point out, it is also pretty efficient in doing so.

Figures: The previous figure is one of the slides. The below figure is an illustration of the "label update" procedure that can be introduced to equip the algorithm with tracking capability. I.e., in a time-varying environment the regression surface can be changing, as indicated by the new data pattern (xn,yn). The proposed procedure updates the labels in the vicinity of this point to reflect those changes.

But enough talking about this paper. My friends and colleagues are also presenting theirs, so if you are at the ICASSP be sure to check them out:

Il Park and Jose Principe, "Quantification of Inter-trial Non-stationarity in Spike Trains from Periodically Stimulated Neural Cultures" (Multivariate and Multimodal Analysis of Brain Signals, Tuesday, March 16, 13:50 - 14:10).
Sohan Seth, Jose C. Principe. A conditional distribution function based approach to design nonparametric tests of independence and conditional independence (Learning Theory and Models III, Wednesday, March 17, 16:00 – 18:00).
David Ramirez, Javier Via, Ignacio Santamaria, Roberto Lopez-Valcarce, Louis L. Scharf, "Multiantenna Spectrum Sensing: Detection Of Spatial Correlation Among Time-Series With Unknown Spectra" (SPCOM-L1: Spectrum Sensing for Cognitive Radio I, Tuesday, March 16, 14:10 - 14:30).
Javier Vía, David Ramírez, Ignacio Santamaría, Luis Vielva, "Widely And Semi-Widely Linear Processing Of Quaternion Vectors" (Detection and Estimation Techniques - II, Thursday, March 18, 13:30 - 15:30).
Luis Vielva, Javier Vía, Jesús Gutiérrez, Óscar González, Jesús Ibáñez, Ignacio Santamaría, "Building A Web Platform For Learning Advanced Digital Communications Using A Mimo Testbed" (Signal Processing Education: Signal Processing Education, Tuesday, March 16, 14:30 - 14:50)
Abhishek Singh, Jose Principe. Kernel width adaptation in information theoretic cost functions (Learning Theory and Models III, Wednesday, March 17, 16:00 – 18:00).
Abhishek Singh, Jose Principe. A closed form recursive solution for maximum correntropy training (Learning Theory and Models III, Time: Wednesday, March 17, 16:00 – 18:00).
Abhishek Singh, Tejaswi Tamminedi, Guy Yosiphon, Anurag Ganguli, Jacob Yadegar. Hidden markov models for modeling blood pressure data to predict acute hypotension (Bioinformatics and Biomedical Signal Processing, Tuesday, March 16, 16:00 – 18:00).

Thanks to Memming for the blog post idea!

PS: The paper ~~will be available in short on my~~ ~~academic web page~~ [update:] can be downloaded from my publications page, and I also plan to release the Matlab code either there or here in the near future. Be sure to ask me about it if you can't find it.

ChromeEye, a TinEye extension for Google Chrome

2009-12-13T23:09:00.010+01:00

I made a simple reverse image search extension for Google Chrome. Reverse what? Reverse image search. Think Google, but instead of text you enter an image. Ah okay. Please continue.

The ChromeEye extension uses the image search engine TinEye, whose technology allows you to find exact copies of images (including scalings and slight distortions).

Version 0.1 of ChromeEye (updated: now latest version) basically mimics the TinEye bookmarklet behavior, i.e. it lists all images of the current page with links to TinEye results. Let's hope Google Chrome allows for editing the browser context menu soon, so you can right click and select a "Search Image on TinEye".

Meanwhile, you can download the latest version of ChromeEye directly from this link, while the complete source code is available on GitHub. So until the oficial TinEye plugin gets released, have fun with this easy solution.

Update 14 December 2009: ChromeEye now has a page on the google chrome extensions site. Version 0.2 was also released, with a nicer logo.

Update 17 June 2011: ChromeEye is now called Reveye and it includes different search engines, such as Google, Tineye and Cydral.

Drawing beautiful block diagrams in LaTeX

2009-12-04T14:15:00.009+01:00

I used to put together the illustrations in my publications with a mixture of Inkscape, CorelDraw and psfrag, but obviously this couldn't continue for a long time without becoming an insult to graphic design.

For my latest writings I am returning to the basics: PSTricks. (I know there are lots of other programs to draw block diagrams, but I didn't find any that allowed enough customization). To avoid starting from zero, I am using the PSTricks Signals and Systems package (pst-sigsys), which is easy to customize if you keep the PSTricks user guide (pdf) at hand.

Here's a simple example:


\documentclass{article}
\usepackage{pst-sigsys}

\pagestyle{empty}

\begin{document}

\begin{figure}[ht]
\centering %
\begin{pspicture}[showgrid=false](0.5,-1.2)(9,1.55)
%--- Define blocks ---
\rput(0.5,0){\rnode{s}{$s[n]$}}
\dotnode[dotstyle=square*,dotscale=0.001](1.7,0){dot}
\psblock(3,.75){H1}{$H_1(z)$}
\psblock(3,-.75){H2}{$H_2(z)$}
\psblock(5.8,.75){B2}{$\hat H_2(z)$}
\psblock(5.8,-.75){B1}{$\hat H_1(z)$}
\pscircleop(7.7,0){ominus}
\rput(9,0){\rnode{e}{$e[n]$}}

%--- Connect blocks ---
\psset{style=Arrow}
\ncline[nodesepA=.15]{-}{s}{dot}
\ncangle[angleA=90,angleB=180]{dot}{H1}
\ncangle[angleA=-90,angleB=180]{dot}{H2}
\ncline{H1}{B2} \naput[npos=.5]{$x_1[n]$}
\ncline{H2}{B1} \naput[npos=.5]{$x_2[n]$}
\ncangle[angleB=90]{B2}{ominus} \naput[npos=.5]{$z_1[n]$}
\ncangle[angleB=-90]{B1}{ominus} \naput[npos=.5]{$z_2[n]$}
\ncline[nodesepB=.15]{ominus}{e}
\end{pspicture}
\end{figure}

\end{document}

which, after some retouching of the .sty file produces this pretty diagram:

Update 6th December 2009: Here's a slightly more complicated diagram, based on the same codes as the above example.

Kindle DX for Scientific Papers?

2009-05-08T08:27:00.006+02:00

Just out of curiosity, I wondered what scientific papers look like on the Kindle DX's 1200 x 824 pixel resolution. The native PDF support is great, but exactly how readable are those tiny little subscripts in mathematical formula?

Although it will probably take some time until I can lay my hands on a real device, I did the test by scaling a page of one of my own papers to the Kindle's resolution. Check for yourself by clicking on the thumbnails below.

mode	original version (vectorial)	scanned version (IEEE archive)
portrait (full page zoom: page fits within 824 x 1200 pixels)
landscape (width rescaled to 1200 pixels)

(The original PDFs can be found here and here.)

In portrait mode, some details might be difficult to distinguish, especially when the PDF is generated from a scanned image. Check for instance this formula, from the scanned version in portrait mode:

But in landscape mode everything looks perfectly readable (at least in this probably very unrealistic experiment). This is the same formula, as it appears in landscape mode:

So I might buy one when it's launched in Europe. But the price needs to be slightly lower. $300 (or its equivalent in €) would be fine. And it needs an SD slot. And Wifi. Apart from that it seems to be fine.

Pictures on ClusterShot

2009-04-24T19:48:00.006+02:00

Time to boost my picture sales!

I am putting some of my pictures for sale on ClusterShot, a new photo marketplace I found this week. Until now I only had my pictures sitting on flickr, and every time somebody wanted to buy one that meant paperwork.

ClusterShot seems to make the process pretty easy, for buyer and seller. And apart from that, their design is very neat. There are still a lot of features missing and issues to work on, but let's give them some time. In the meantime, I invite you to go check out my page and haggle on my pictures: clustershot.com/steven2358

Illustration: some Spanish lolcows

The Love Injections web sites

2009-02-17T18:42:00.000+01:00

Now that Valentine is over, can we please proceed with the serious matter?

I programmed the websites for the release of the Love Injections book, written by Kim Brusselmans. Here are the English version (Love Injections) and the Spanish version (Inyecciones de Amor). This book might be described as "an original attempt to inspire people to boost their relationships", but to get a better idea of what that means you should check this page: Example pages.

The book is on sale in a few shops in Spain and Belgium, and after the crazily successful book presentations in Barcelona and Ghent, we also opened online shops for both countries.

The sites were done in Drupal, with a fairly simple Ubercart setup to deal with the shopping part. The minimalist graphic design was done by Luxoa and programmed as a subtheme of the Zen theme.

The silent MP3: To bed with the iPod Shuffle or Zen Stone

2008-03-11T18:56:00.003+01:00

Flaws in products call for hacks, so that's why I tried to solve the following annoyingness.

Lots of people listen to their mp3 player in bed and fall asleep with the music still on. "Shuffle" players like the iPod Shuffle and Zen Stone don't have a "do not repeat" option, so they just continue playing songs until the battery dies. That will most probably disturb your good night's rest, and apart from that you will wake up with an empty battery. And we don't want either of these, do we.

I figured that a small hack could solve both problems: put a very long silent MP3 in your playlist. You could peacefully sleep over it and it wouldn't drain your player's power either. So I generated a set of MP3 files that occupy just 3.5 MBytes per hour of true quality silence. They are encoded at 8 bits, mono and 8 KHz (the minimum sound quality playable by most players). The links are at the bottom of this post, packed as .rar archives. This is how it works:

Determine how long you want to sleep, e.g. 7 hours.
Download the corresponding silent mp3 file and put it in your playlist just after the track by which you will likely fall asleep. If you don't have a playlist, simply rename the mp3 so that it is played after the last music track you want.
Go to bed with your favorite music, fall asleep and enjoy the silence.
Get up, switch off your mp3 player, and have a croissant.

Suggestions are open for other uses of these silent MP3s. So onto the next problem now: avoiding being strangled by the headphone chord while sleeping.

Files (compressed with RAR):

sleeper1.mp3 (1 hour of silence: 3.5 MB unpacked, 2.83 KB packed)
sleeper2.mp3 (2 hours of silence)
sleeper3.mp3 (3 hours of silence)
sleeper4.mp3 (4 hours of silence)
sleeper5.mp3 (5 hours of silence)
sleeper6.mp3 (6 hours of silence)
sleeper7.mp3 (7 hours of silence)
sleeper8.mp3 (8 hours of silence: 28.1 MB unpacked, 17.7 KB packed)

Update 16th April 2009: Due to popular demand I'm adding a complementary set of shorter MP3s, for all you sleephackers out there. This should cover your napping needs.

sleeper.01.mp3 (1 minute of silence: 58.9 KB unpacked, 239 B packed)
sleeper.03.mp3 (3 minutes of silence)
sleeper.10.mp3 (10 minutes of silence)
sleeper.30.mp3 (30 minutes of silence: 1.7 MB unpacked, 1.26 KB packed)

Hosting Uptime/Downtime Calculation

2008-02-18T01:27:00.008+01:00

While searching for hosting companies for a number of websites I'm working on, I noticed that a lot companies in Spain try to sell 99.5% guaranteed uptime as being very little downtime. Since this seemed like less than the average to me, I figured I might do a little calculation.

guaranteed uptime	max downtime	downtime / week	downtime / month	downtime / year
99%	1%	1.68 hours	7.2 hours	3.65 days
99.5%	0.5%	50.4 minutes	3.6 hours	1.825 days
99.9%	0.1%	10.08 minutes	43.2 minutes	8.76 hours
99.99%	0.01%	1.008 minutes	4.32 minutes	52.56 minutes
99.999%	0.001%	6.048 seconds	25.92 seconds	5.256 minutes
99.9999%	0.0001%	0.6048 seconds	2.592 seconds	31.536 seconds
99.99999%	0.00001%	0.06048 seconds	0.2592 seconds	3.1536 seconds

While most of these scenarios are not really realistic, it is interesting to see what companies are really offering. Reading about the recent Amazon S3 outage makes one want to place things into context. Amazon's goal is to offer 99.9% uptime, which means a maximum of 0.1% downtime and therefore not more than 43.2 minutes of downtime per month. The outage took little more than 2 hours which qualifies it as a 99.5% host this month. That's a long way from the 99.999% they should be offering for cloud computing indeed.

Photo tricks pt.2: Fun with Escher and Droste

2007-10-30T15:51:00.001+01:00

un excès de bricolage
Originally uploaded by Steven2358

Last weekend I dropped some of my pictures in the weirdly whirling image manipulation device that is the Droste effect. It looks fun, it is fun, and it isn't that hard to achieve at all. Just follow the steps in this fine Droste effect tutorial and get confused faces in no time.

Spring is in the Air - I'm finishing some websites

2007-05-01T09:05:00.000+02:00

AdventureImagery.com, the online shop for extreme sports and related stock footage. Check out the crazy videos! I particularly like the skydiving, kite surfing and bicycling footage.

Santander-Unico.com, the "quality guide to Santander". Indispensable for anybody visiting Santander (Spain). Includes some google maps and the nice shadow hack tastypopsicle made of the onion skinned drop shadows.

AulaDeOratoria.com, the first Spanish portal for everybody interested in getting his message to his audience.

La Ring 01

2007-01-12T21:01:00.000+01:00

(... or how to destroy your inner ear cells)

Ever noticed how difficult it is to hear your mobile phone in a noisy place like a sports stadium or crowded bar? "That's because there's a lot of noise" you think. But there's an additional factor: polyphonic ringtones are a lot harder to hear than the good old monophonic ones.

The problem with polyphonic ringtones is that since they divide their output power over the complete audible spectrum, they get easier masked by environmental noise. Monotonic ringtones stand out, putting all their power in a single frequency at a time. Anyway, polyphony is here to stay, so let's make the best out of it.

Here's how to make a ringtone that you can hear well in noisy situations (requirements: audio editor software + mobile phone that can play mp3s):

Take your favorite song that has a lot of high frequencies (audio fx, percussion, hiss noises,...) and chop it up a bit. I took Vitalic's "La Rock 01" but you might as well try out some of Dave Tipper's stuff.
Remove bass frequencies and add treble, e.g. with a filter like this:
Transfer it to your phone, sit back and wait until you get called.

My ringtone is available right here, ready for download and transfer to a mobile phone. You can preview it with the applet below, although it sounds different on a pc speaker than on a phone. And be warned: listening to this kind of stuff a lot is not good for your ears.

Tags: mp3, cell phone, audio, diy

Wedders with Hollywood ambitions

2007-01-05T02:18:00.000+01:00

You like movies and you're about to get married? Enter the fine initiative of the people of bodametraje.com, who cast the pair of you in the roles of real movie stars, making a complete short movie with you as the stars/ heroes/ villains/ whatever suits the script. The idea was launched in Spain by bodametraje.com, and the wedding guests really seem to appreciate the dvds with the short films (which are mostly handed out at the wedding). Or, to quote the production team:

"Bodametraje.com ofrece la posibilidad de dar un toque inovador y exclusivo a su boda mediante la realización de un cortometraje de cine protagonizado por los novios, un "bodametraje".

Usted elige un tema y la organización se encarga de escribir el guión. El tiempo de rodaje suele ser de tres días aproximadamente, y no es necesario saber actuar, un buen trabajo de postproducción conseguirá un resultado espectacular.

El cortometraje podrá ser proyectado en pantalla gigante durante el cocktail de recepción o en algún momento del banquete o bien se repartirá un DVD con el mismos a cada invitado como recuerdo de los novios sustituyendo los típicos puros, bombones, o figuritas... (que tampoco nos sorprenden).

Sus familiares y amigos quedarán impresionados al descubrir su lado más Bond... y les agradecerán la originalidad, pues quién quiere ver otra boda estándar?

No solo tendrá el corto sino que además podrá disfrutar de las tomas falsas y entrevistas tipo Hollywood. El DVD tendrá un total de 30 minutos."

I programmed the flash website, easy and to the point. Check it, there's a nice shortfilm under the "ofrecemos" menu.

Tags: hollywood, wedding, short film, boda, cortometraje

Some more ActionScript programming

2006-08-25T20:05:00.001+02:00

Last week I finished the flash programming of two design websites, one for prestigious designer José Piris and another one for the stylish interior design shop Cromoduro. All was done in ActionScript, with a new library of prototype functions I made for flash. The design itself was done by Kim B., and the result is smooth enough - another one for the archive.

Tags: design, diseño, interior, flash

Let's mash up some Google

2006-08-25T19:30:00.000+02:00

I've been digging a little into the google maps api, but apparently you don't need to know anything to do some things. Thanks to Google maps mania and Wikimapia, I got the frame below (in 49 seconds), with a nice tagged Google map of Santander. Pretty cool indeed.

Photo tricks pt.1

2006-06-20T00:21:00.000+02:00

I just learned how to finetune Autostitch to make some pretty nifty panorama pictures. Check my last one, stitched together from 11 pictures. I also did some post-processing, but that might need some more finetuning. Anyhow, this looks like a nice way to toy with my weekend pictures.

Tags: autostitch, photo, flickr

Speech Analyzer v1.0

2006-06-19T23:50:00.000+02:00

Phew, I'm starting to finish off some side projects. Yesterday I made the finishing touches on a speech analyzer program meant as an accompanying tool for a set of oratory courses in Spanish. It's programmed in .NET and makes some beautiful real-time plots of voice frequency (pitch) and speech rate. Some technical details: real-time pitch estimation was done by Noll's algorithm, real-time speech rate estimation was a mix between enrate and a newer correlation-based method. For the signal processing part we relied on the Exocortex.DSP FFT library, and for the graphics we used the ZedGraph classes.

Good. One for the archive.

Tags: dsp, speech, pitch, real-time, programming, .NET

Taking the colour Sudoku to the previous level

2006-01-12T16:27:00.000+01:00

I received a few requests for a version of the classic sudoku, after releasing a flash version of the colour sudoku. Since that meant changing just a few lines of code, I made it in a moment yesterday. So here it is, a normal sudoku puzzle in flash!

As with the colour version there are 2 difficulty levels, and a puzzle number can be chosen of the 1.000.000.000.000.000 possibilities or it can just be set to generate random puzzles. Comments are welcome, just post them below here, or on the feedback page of the puzzle.

Next up should be a high-score table, but I'm busy with other stuff at the moment.

Tags: sudoku, puzzle, flash

Google PageRank Calculator

2006-01-08T16:30:00.000+01:00

Since I don't always have a Google toolbar on the computers I work on, and I often want to know the PageRank of a site, I made a little PageRank calculator tool that does just this and nothing else. To use it, fill in an url and press the button.

Have fun.

Tags: google, pagerank, web, utilities

Taking Sudoku to the next level

2005-10-04T18:50:00.000+02:00

Last week I released a puzzle on squobble.com, conveniently dubbed the "Squobble Puzzle". It's quite similar to Sudoku, so the hardcore Sudoku players might find this brain-teaser an entertaining alternative.

Basic Sudoku is played with 9 symbols, which are generally the numbers 1 till 9. Other than numbers, these can be anything from 9 letters to 9 colours (for the rules, check out Wikipedia). I combined a "number Sudoku" with a "colour Sudoku" to obtain the Squobble Puzzle. This has some interesting consequences. For instance, only certain combinations of numbers and colours are possible. Also, the classic Sudoku tricks can be applied on numbers, colours or both, and even new solution methods can be found.

I included a beginner and expert level (if you're good they should take you about 5 and 15 minutes respectively), and the standard options: reset, choosing a puzzle by number or randomly. Moreover, every puzzle has a unique solution. Finally, to guarantee your satisfaction, dear reader, all of this was molded neatly into a Flash puzzle with drag-and-drop support.

Anyway, the list of features-to-add includes a charts system for best times and other records, a classic Sudoku page, etc. Whether I implement these will depend mostly on visitor feedback and the local weather forecast (a lot of rain means more features).

Enjoy!

Tags: sudoku, puzzle, flash, color, colour

Plaza 63

2005-07-02T19:01:00.000+02:00

I made a "tourist guy" appearance in this short movie (and part of the soundtrack as well). The acting was a bit too improvised to post a link to the online version, however - I have to preserve at least some cool.

Dromedarios Inc.

2005-06-21T18:30:00.000+02:00

I made this ad for the Spanish deepline.net, pretty hilarious in terms of combinations of jokes if you ask me. The funny thing is, some people didn't get it.

Movistar Art #1

2005-05-03T17:18:00.000+02:00

Movistar España, the leading mobile phone operator in Spain, has launched a 75 million euro campaign to promote its new logo, a stylized letter "m". Since its start one month ago, the initiative has drawn huge attention from the public, mainly because of the air of mystery it was initially surrounded by.

Nevertheless, what was once new quickly got boring, and since the new image didn't really seem to boost sales, a new approach was taken this week in an attempt to make a lasting impression on the potential customer.

As reported by the Spanish national police, street violence has dramatically risen since the new posters started appearing everywhere. Soon more on this subject.

[edit] Apparently some well known Spanish actress also participated in the campaign.

Matlab Art #2

2005-04-05T17:36:00.000+02:00

Origin of the image: 1 frame of the evolution the output of a multi layer perceptron is going through while it tries to adjust the different curves to straight lines. In this case the adjustment was obviously heading towards an explosion so I stopped it at the time of the image.

Shazam@home

2005-04-05T16:51:00.000+02:00

Anyone who has been using the music tagging service by Shazam might have wondered how on earth their computers are able to recognize audio that fast and under such extremely noisy conditions. The solution, as I found out after a quite short internet search, is a pretty smart algorithm developed by Dr. Avery Wang. Thanks to the hints he gave in a presentation at the ISMIR2003 conference, I was able to program the whole thing in Matlab in one evening.

First I did a test with a few songs on my pc. After the program trains itself with the songs, it takes a 5-second fragment of one of the songs, highly degrades it with noise and then tries to recognize it. It turned out to work just fine, and above all, very fast. Then I went for the heavier version, and fed 1000 different audio tracks into the training part of the program. To my surprise, it continued to recognize the music perfectly, in much cases up to noise degradation levels at which the human ear isn't capable anymore of recognizing it, which was nice.

That being written, if you want to get the Matlab code to play around a bit at home with, just contact me.

Matlab Art #1

2005-03-25T18:52:00.000+01:00

"Oh no! This simulation's gone terribly wrong! Just look at the graph, it's useless!" In that case, let's consider it art :-) So this is one of those images that look like, but weren't supposed to be art.

Origin of the image: real parts of the poles of the filters representing two fragments of mixed voices.

Let's eat some squobbles!

2005-03-07T19:04:00.000+01:00

With the transition from Squobble v1.x to Squobble v2.0 in mind, I started the automatic deletion of the links in Squobble today.

Previously, there were always 150 links available on the board, and the least popular ones were replaced with new suggestions. Since I want them all out of the board before switching to the new version, I changed a parameter so that from today on it will "degenerate" automatically. Every day a few of these 150 will be eaten by the archive, until there are none left. Let's hope v2.0 will be ready at that moment.

Getting into ActionScript

2005-02-28T09:38:00.000+01:00

I learned Flash ActionScript last month, to do the special effects of some design pages. We applied them to one page and basically recycled them for the other two (here and here - Spanish warning!). They're all done now, and I think they're quite esthetically pleasing;-) If you're into ActionScript, check the the dynamic menu, the text effects and the flashy image transitions. Pretty basic I suppose, but not a bad start either.

Here we we go then!

2005-02-24T16:58:00.000+01:00

Okay, it was about time I started this blog, common sense just forces me to. I mainly plan to post info and comments on a few things I'm doing, and obviously, feedback is welcome (since it is the first step towards improvement).

First of all there's my wicked internet experiment, Squobble. I won't explain on it too much now, since it's a pretty simple thing to understand. Moreover, its behaviour is still changing, so I will just comment the parts as they stabilize. In short, it's pure fun, leaning over to the crazy side. Then, there's the audio related investigation stuff I do, ranging from voice recognition to music processing algorithms. Pure fun too, but at least this matter is scientifically justified. And last and least there's a tiny bit on graphic design (mostly web design) just for the love of it. Anyway, a nice "Welcome to you, dear reader!" seems to be appropriate to finish this first post.

PS1: On the to do list: change the crappy template design of this page in something a little more neat.
PS2: Welcome to you, dear reader!