Re: [PD] OCR with Puredata?
On 6/1/11 8:54 PM, Mathieu Bouchard wrote: On Wed, 1 Jun 2011, Alvin O'Sullivan wrote: It is in the early concept stages, but I was thinking that the user interface would be a piece of glass/plastic the input device would be a webcam on the inside of a box/resonator. The ambient light outside of the box (no pun intended) would be enough of a difference to read the opaque markings as parse-able characters. Ok, this is two-dimensional reading, which is already a lot more ordinary. If characters are upright (non-rotated) it may be even more ordinary. This means it'll be easier to find existing support for reading characters, or to (re)invent that support. Beyond that, I don't know much more about OCR. ___ | Mathieu Bouchard tél: +1.514.383.3801 Villeray, Montréal, QC I've not used the neural net libraries/abstracts before so I will give it a try and see how it goes. It sounds as though having a good grid size to input method resolution would be the first step. Some of the image processing can be handled by the web-camera driver: contrast, flipping the image etc. The interface will be part of a sculpture that act as resonator, so depending on the input I can catch and process my patch will respond with auditory feedback. The key is to keep the user interested and willing to play with this thing. Marks or gestures would be enough if character processing were not practical. I will also have to some how recorded the different session states later to be recalled when a user is not available. Thanks again for all of your help. A ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
Re: [PD] OCR with Puredata?
Le mercredi 01 juin 2011 à 12:51 -0400, Tedb0t a écrit : Is it just me, or it sounds like it's going to take a lot of preprocessing before you can even think of feeding it to a neural network ? Black/white thresholding and resolution reduction, that's it. I also use grey images and it was working fine. Human vision is made of a lot more layers of neurons than we can hope to deal with in artificial networks. It sure is. Luckily you only need a few to do some basic OCR. You could try a grid of something like 10x10 pixels just to start, which would require 100 input neurons. The lower the resolution, the higher your error rate will be, but you can find good compromises. I was working with a grid of 16x16 pixels and it was OK for me. At what angles should characters be recognised ? Well, that's the beautiful thing about neural nets—it just depends on how you train the net. If you want the net to be able to recognize tilted letters, you can add tilted letters to the training sets. It can affect Yep, if you train your NN with angles, deformations, etc. You should get good results. ++ Jack But making an OCR using ANN is a lot lot more work than using an OCR library. Making a Pd-to-OCR-library interface is less work than making an OCR abstraction library Agreed, strongly. OCR is a really excellent exercise in neural nets (look up Self-Organizing Maps or Kohonen networks), but it's a lot of work. It would be faster by far to set up an interface as Mathieu suggests. However, if you want to go through with it anyway, I'd love to help! It looks like the defacto open ocr lib is Tesseract: http://code.google.com/p/tesseract-ocr/ This would be great to have in Pd. Incidentally, I can't see any reason why Pd would be bad at it, since the ANN external uses the C FANN library, which is likely what any other library would use in the first place. ±tedb0t On Jun 1, 2011, at 11:49 AM, Mathieu Bouchard wrote: On Wed, 1 Jun 2011, Jack wrote: You can do this with the use of artificial neural network (for character recognition). There are externals for Pd : http://pure-data.svn.sourceforge.net/viewvc/pure-data/trunk/externals/ann/ Is it just me, or it sounds like it's going to take a lot of preprocessing before you can even think of feeding it to a neural network ? Human vision is made of a lot more layers of neurons than we can hope to deal with in artificial networks. At what angles should characters be recognised ? Which colour on which colour ? You better settle those things first, so that you can figure out how you can reduce your data beforehand. But making an OCR using ANN is a lot lot more work than using an OCR library. Making a Pd-to-OCR-library interface is less work than making an OCR abstraction library... and it isn't necessarily because Pd would be bad at it (I don't know about that). It's more because it takes a lot of knowledge to make an OCR library from nearly scratch. ___ | Mathieu Bouchard tél: +1.514.383.3801 Villeray, Montréal, QC ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list signature.asc Description: This is a digitally signed message part ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
Re: [PD] Problems with FLOSS Manual on Scribd
Probably should've searched before posting to you all.. the old FLOSS page is indeed up. Sorry, -Ben On Thu, Jun 2, 2011 at 10:37 AM, Ben Baker-Smith bbakersm...@gmail.com wrote: I thought Scribd was gonna be cool, but now the pages don't load and it wants me to pay $5 to download the pdf, which I would have to do in order to read the manual since it's not loading properly. Lame. Is the old FLOSS site still available? -Ben ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
[PD] Problems with FLOSS Manual on Scribd
I thought Scribd was gonna be cool, but now the pages don't load and it wants me to pay $5 to download the pdf, which I would have to do in order to read the manual since it's not loading properly. Lame. Is the old FLOSS site still available? -Ben ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
Re: [PD] OCR with Puredata?
On Wed, 1 Jun 2011, Tedb0t wrote: Well, that's the beautiful thing about neural nets—it just depends on how you train the net. If you want the net to be able to recognize tilted letters, you can add tilted letters to the training sets. That would take a damn lot more elements in the training set ! It looks like the defacto open ocr lib is Tesseract: http://code.google.com/p/tesseract-ocr/ This would be great to have in Pd. What a name. I mean, it's geometry-related, but Tesseracts are about four-dimensional spaces. :} Incidentally, I can't see any reason why Pd would be bad at it, since the ANN external uses the C FANN library, which is likely what any other library would use in the first place. Reasons why a language would be bad at it are obviously because of what would be different, and not what would be the same. FANN would be the same, and the rest would be different. BTW, [ann_som] is not using FANN, did you know that ? ___ | Mathieu Bouchard tél: +1.514.383.3801 Villeray, Montréal, QC ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
Re: [PD] CVs
On Mon, 23 May 2011, Simon Wise wrote: Basically I am interested in the notion that we could recognise groups of the same size having in some way the same pattern, without going on to map these patterns onto a series of numbers. It certainly is useful to map these patterns to numbers, but all the same they are recognisable simply as patterns. Two things together seemed interesting in this regard. If there's some part of brain lobe that can recognise patterns made of N similar things, those patterns are usually called numbers anyway. That's merely a difference of terminology (but differences of terminology do matter a lot). First the ability of some people to recognise quite large groups directly, without counting. The description of this process did seem to suggest that it was something other than clever, quick shortcuts to counting ... there was quite a lot involved because that was an obvious possibility and the discussions and tests led the researcher to conclude that it was not done this way. I hope that it's better than what survey houses do. They ask people « which political leader do you prefer ? » and then they report « 471 people out of 1003 prefer XYZ » rather than « 471 out of 1003 people claim that they believe that they prefer XYZ ». You'd like to think that scientists would want to accurately report their findings... and not look like they are gullible vis-à-vis their respondents. Next morning, anyway, major newspaper conglomerate headlines « 532 out of 1003 people reject XYZ », which is even more false. Anyway... I haven't read Sacks. Certainly the languages would have been near extinct, more complex ideas are useful often, and it is probably easier to learn a language that has the vocabulary to expresses them than invent a new vocabulary and syntax to add to an old language. That sounds like a colonialist perspective. You have no idea how easy it is for people to add words to their own language. The hard part is to decide to do it. The examples I recall described were not about a disgust for numbers ... Replace disgust by something equivalent such as the impression that a life without any numbers is more noble or authentic and that numbers are superfluous concepts brought by foreigners for no good reason. perhaps it was just they had found no need to communicate the idea of numbers, how about a reasoning like such and such requires numbers, therefore we don't need it ? How about that those are the numbers that you can't possibly do without even if you wished very strongly to not use « numbers » ? I'm wondering more about how these things can be described other than mapping to numbers, since - to pull back to Pd - we often do the opposite in computers, and map an unordered set to a series of integers just because it is convenient to deal with integers, eg passing messages around in lists (which are still ordered, even if the order is meaningless except by convention, and accessed by their integer index). When I pass « this is a sentence » as a plain list, $4 = sentence, and without doubt, sentence is the 4th word of the sentence, but it doesn't mean that I thought about « 4 » when saying « this is a sentence » : the number only has to be inferred from the data that « this » is the first word, that the next one is the 2nd, the next one is 3rd, and next one is 4th. 4th only means next of next of next of first... in other words, s(s(s(s(0. Words are naturally ordered because they have to be said one after the other, in time, and time is a totally-ordered dimension at that scale. Numbering is very useful in practice, but it is interesting to consider what can be done without it. is 1,549,364 anything other than word in the language of mathematics? well, it's also the sum of squares of 292 and of 1210... ;) That is neat, it was derived as a string of the first digits my fingers hit on the keyboard. So its square root (probably an irrational number) is the length of the diagonal of a rectangular piece of paper with sides 292 1210. Assuming of course that our space is actually Euclidian. Numbers do have lots of nice properties. Yes, the square root is an irrational number. However, you can rotate that square root in the complex plane, to get to (292+1210i), which is rational (and integer). The rotation angle is arctan(292/1210). ___ | Mathieu Bouchard tél: +1.514.383.3801 Villeray, Montréal, QC___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
Re: [PD] CVs
On Thu, 2 Jun 2011, Mathieu Bouchard wrote: Yes, the square root is an irrational number. However, you can rotate that square root in the complex plane, to get to (292+1210i), which is rational (and integer). The rotation angle is arctan(292/1210). or rather, arctan(1210/292)... ___ | Mathieu Bouchard tél: +1.514.383.3801 Villeray, Montréal, QC ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list
[PD] [PD-announce] Pd-berlin meeting next tuesday, 7th June
Hello, next tuesday, 7th June, will be the next meeting of Pure Data users in Berlin at NK (http://www.nkprojekt.de/) - Elsenstr. 52, 2HH 2Etage. For more information, look up http://puredata.info/community/organization/pd-berlin/pd-berlin-users-group. We also encourage you to take an active part, and put up suggestions for topics you want to talk about / topics you want to be talked about. Pd-Berlin Google group: You can join the open group http://groups.google.com/group/pd_berlin, and make questions to the users there. Doors are open from 20h-20h15. After that they'll be closed, and you will have to call someone from the Pd-meeting to get in. To get a telephone number to call or confirm assistance you can write to info_at_minitronics.net. Please, don´t call to the staff of NK to open the doors. They let us use the space but we have to take care about having the meeting without producing any disturbance to them, and to clean the space after the meeting. We would apreciate if you would send us a small mail to info_at_minitronics.net with your name, Pd experience and interests, so that we know how many people might be coming. Or put your name in the pd-berlin wiki page. We would like to thank the support and willingness of NK in the organization of these events. João Pais ___ Pd-announce mailing list pd-annou...@iem.at http://lists.puredata.info/listinfo/pd-announce ___ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management - http://lists.puredata.info/listinfo/pd-list