Search found 22 matches

by BruceRobertson
June 20th, 2014, 11:17 am
Forum: Polytonic Greek OCR
Topic: Improving Polytonic Greek OCR
Replies: 18
Views: 22457

Re: Improving Polytonic Greek OCR

Hi, Seumas -- I take it these letters would be in PG 29-32? I can process them, to the best that our scans and ability allow, in the next couple of weeks. However, for this specific project you also might want to take a look at the Loebs we've already processed, available in this list: http://heml.m...
by BruceRobertson
May 31st, 2014, 9:54 am
Forum: Polytonic Greek OCR
Topic: Ciaconna: Resources for Greek OCR with Ocropus
Replies: 3
Views: 11747

Re: Ciaconna: Resources for Greek OCR with Ocropus

The parallelization code assumes a Sun Grid Engine scheduler and 20 - 40 cores with shared memory, since that's what I had last year. The Gamera code requires many standard Python libraries. But as I say, for many uses the Ocropus-based 'Ciaconna' library is as good or better, and it doesn't really ...
by BruceRobertson
May 25th, 2014, 7:03 pm
Forum: Polytonic Greek OCR
Topic: Assignment 3: PG 94 col. 1593-4
Replies: 17
Views: 22214

Re: Assignment 3: PG 94 col. 1593-4

Do you mean a list of words / forms that you apply the Levenshtein distance process to? Or is it like, a list of requiring / needs / attracts - required things? E.g. κατά requires something in the accusative, a verb attracts an adverb (weaker than require), nouns in the accusative might be the obje...
by BruceRobertson
May 25th, 2014, 5:16 pm
Forum: Polytonic Greek OCR
Topic: Primary vs. Secondary Resources
Replies: 7
Views: 12676

Re: Primary vs. Secondary Resources

I'm happy to recognize anything anyone wants, using existing classifiers, so requests on that level are quite welcomed. We do, however, want to focus our training labour here on just a couple of texts. Sophocles we have already run from an archive.org copy. Here is the OCR output that we have: http:...
by BruceRobertson
May 24th, 2014, 5:45 pm
Forum: Polytonic Greek OCR
Topic: Assignment 3: PG 94 col. 1593-4
Replies: 17
Views: 22214

Re: Assignment 3: PG 94 col. 1593-4

In 1593 line 49 Φαρισαῖοι
This is the correct form (with circumflex). Migne was cheap with the ink, so sometimes letters and accents are not fully inked in. This is why his volumes are black-belt OCR :-)
by BruceRobertson
May 24th, 2014, 5:36 pm
Forum: Polytonic Greek OCR
Topic: Assignment 3: PG 94 col. 1593-4
Replies: 17
Views: 22214

Re: Assignment 3: PG 94 col. 1593-4

Thanks for all these questions and ideas. What you propose known as applying a 'language model' to OCR output in order to correct the output so it conforms to that model. We do have a spellcheck program that uses Levenshtein distances and a large (3million) dictionary to correct some of the sorts of...
by BruceRobertson
May 24th, 2014, 1:46 pm
Forum: Polytonic Greek OCR
Topic: Assignment 2: PG 93 col. 1571-2
Replies: 11
Views: 18368

Re: Assignment 2: PG 93 col. 1571-2

Here's a sort of rule-of-thumb: assuming we'll never get up to 100%, each doubling of ground truth gets us twice as close to, say 98% character accuracy. So if we're at 93% now, ten more pages might get us to 95%. I think ultimately, given how large and diverse Migne is, I'd like to have about 50 pa...
by BruceRobertson
May 24th, 2014, 11:59 am
Forum: Polytonic Greek OCR
Topic: Ciaconna: Resources for Greek OCR with Ocropus
Replies: 3
Views: 11747

Ciaconna: Resources for Greek OCR with Ocropus

Most of the results at http://heml.mta.ca/lace are from the Gamera-based Rigaudon engine, which requires a parallel computing environment. However, recently, I've been using that data to train a much simpler engine, Ocropus 0.7. If you have a Unix environment and are comfortable installing software,...
by BruceRobertson
May 24th, 2014, 8:36 am
Forum: Polytonic Greek OCR
Topic: Primary vs. Secondary Resources
Replies: 7
Views: 12676

Re: Primary vs. Secondary Resources

Thanks for these suggestions!
by BruceRobertson
May 23rd, 2014, 10:03 am
Forum: Polytonic Greek OCR
Topic: Assignment 3: PG 94 col. 1593-4
Replies: 17
Views: 22214

Re: Assignment 3: PG 94 col. 1593-4

Feel free to provide corrected Greek and leave the Latin as is. Someone else can do that.

Go to advanced search