Tuesday, December 27, 2011

Driverless cars

Remember that guest lecture about pedestrian detection? Their most impressive feat was that the system could temporarily take control of the car and perform a manoeuvre to avoid pedestrians. Well, let's just say that at Stanford they do things a little more advanced.

 

To be fair, these driverless cars employ protruding sensors that are way more obstructive than the ones used by Daimler. In addition, the roads on the other side of the pond are easier to navigate for cars than those in Europe. Still, pretty impressive, huh?

For more information, here's the website of Sebastian Thrun, director of the Stanford AI lab and lead behind this project.

Monday, December 26, 2011

Registering for courses

Here's another reminder that everyone needs to register via SIS for courses for the entire second semester.

I was also wondering if anyone else is able to register for the profile project of the track gaming (in period 2c)? Its the only profile project I cannot find... but thats the one I want to register for.

Saturday, December 24, 2011

ELPL Assignment 2

So yeah here goes, assignment 2:

TijmenFusRo(brecht)DahKoen

Seriously, 2000 words is not even nearly enough to go in-depth into the articles. We had to suffice with giving a rather general comparison of the models :(. How did you guys do? Had the same feeling?

Thursday, December 22, 2011

Tracker videos

Extracting frames from a video can be quite a hassle in matlab. You should use ffmpeg instead.

You can install it on ubuntu/debian using the terminal
sudo apt-get install ffmpeg

the terminal command for ffmpeg

ffmpeg -i YOURVIDEO -y -an -sameq -r 15 Frame%04d.png

This should the frames in the same format as the soccer video.

I also activated the youtube account for maaiuva@gmail.com so we an upload our tracker videos there :)

Help needed for making a procedural music game


The time for Project AI is swiftly approaching and I would like to do a custom project instead of the pre-defined ones. To make that happen, two things are needed:
  • A fellow student willing to team up with me to work on this project.
  • The approval of one professor who will supervise the effort. 

What is this project about? It's a procedurally generated music game. So, you give it a music track as input and it generates a game level that is in sync with the music. While playing, your interaction with the game is in accordance to the music. The more you follow the music, the better.

More information after the jump.




Sunday, December 18, 2011

Matte and shiny

I am still a bit confused about the matte and glossy surface stuff for the exam. What exactly do we need to know about these models and how do they differ?
Hello.
I guess it's really too late, but if somebody has a few minutes to spare, I would appreciate some guidance as how the following things work and what results do they actually produce:

trec_eval (ok run by itself)
plotRunDifference.pl (run as perl plotRunDifference.pl)
ProcessPR.pl (run as perl ProcessPR.pl)

Thank you.
PS. Yes, it's for the resit assignment, due tonight :)

Wednesday, December 14, 2011

IMS tutorial solutions

Hi guys!

Here I link the solutions for the 1st IMS tutorial, as I solved it, with the great help of Mihai:

Werkcollege 1
Werkcollege 2 (partial)
Werkcollege 3

I hope to transcribe the others during the following days. (almost) finished :-)

Bye!
Dario

Tuesday, December 13, 2011

Do we even need machine translation?

Massive-scale online collaboration is a TED talk given by  Luis von Ahn, known as one of the first people who worked with CAPTCHAs and as the inventor of reCAPTCHA. He argues that machine translation isn't good enough and proposes a crowd-sourcing solution for breaking down language barriers. Interesting stuff!




Here's the website of Duolingo, the translation crowd-sourcing project that was described in the talk. Note that this isn't necessarily antagonistic to our AI approach; the output of this project can be used to provide us with the sentence pairs we so desperately need to train our translation algorithms.

Sunday, December 11, 2011

Quote

No amount of experimentation can ever prove me right; a single experiment can prove me wrong - Albert Einstein

Tuesday, December 6, 2011

The face of a liar

If you liked today's lecture on facial expressions, you'll also probably like this Radiolab episode on deception. Part of the episode is focused on micro facial expressions, which can be used to detect someone wearing a fake facial expression. Paul Ekman, the guy the lecturer mentioned, is featured. (skip the first three minutes if you don't want to hear about a type of snake that plays dead when in danger)

Monday, December 5, 2011

CYK-parser performance

Hi all,

so I ran CYK with Viterbi without smoothing out unknown words and as a result I got:

-- len<=15 --
Number of sentence = 603
Number of Error sentence = 0
Number of Skip sentence = 0
Number of Valid sentence = 603
Bracketing Recall = 31.71
Bracketing Precision = 41.81
Bracketing FMeasure = 36.06
Complete match = 0.66
Average crossing = 1.42
No crossing = 51.41
2 or less crossing = 75.79
Tagging accuracy = 83.71

does this match your findings so far? I want to know before I start smoothing :)

Update:
-- All --
Number of sentence = 603
Number of Error sentence = 0
Number of Skip sentence = 0
Number of Valid sentence = 603
Bracketing Recall = 77.30
Bracketing Precision = 78.17
Bracketing FMeasure = 77.73
Complete match = 21.56
Average crossing = 0.73
No crossing = 70.48
2 or less crossing = 89.39
Tagging accuracy = 92.66

Friendly Reminder

I just want to give a reminder that the registration for the second semester is open and that you should register the courses that you want to do :)

and while we are on topic: any tips on good free choice courses?