Commit graph

583 commits

Author SHA1 Message Date
Robert Sachunsky
6fe02df973 do_image_rotation: fix f93fa12 (do return results) 2024-12-09 16:35:31 +00:00
Robert Sachunsky
d68017037c do_prediction: trigger GC to avoid CUDA OOM 2024-12-09 11:27:11 +00:00
Robert Sachunsky
ad748d0039 do_prediction: avoid code duplication 2024-12-09 10:55:41 +00:00
Robert Sachunsky
c3163caefd avoid indentation 2024-12-05 14:28:17 +00:00
Robert Sachunsky
055463d23a avoid indentation 2024-12-05 09:43:30 +00:00
Robert Sachunsky
aaea2ef463 simplify 2024-12-05 09:40:02 +00:00
Robert Sachunsky
3d88b207fc run: log instead of print 2024-12-05 09:39:55 +00:00
Robert Sachunsky
a520bd1f77 wrap extremely long lines 2024-12-04 23:04:51 +00:00
Robert Sachunsky
cd4e426977 avoid indentation (skip_layout_and_reading_order) 2024-12-04 23:04:48 +00:00
Robert Sachunsky
5b82320707 avoid indentation 2024-12-04 22:09:32 +00:00
Robert Sachunsky
9f12fa241d log-level: only set 'eynollah' logger level 2024-12-04 22:09:15 +00:00
Robert Sachunsky
14beb46224 simplify loading models w/o dir_in mode 2024-12-04 21:07:26 +00:00
Robert Sachunsky
329fac23f6 do not reload enhancement model in dir_in mode, simplify 2024-12-04 18:29:49 +00:00
Robert Sachunsky
3b9a29bc5c simplify dir_in conditionals 2024-12-04 18:19:54 +00:00
Robert Sachunsky
7ae64f3717 RO model: do not reload when in dir_in mode 2024-12-04 16:18:35 +00:00
Robert Sachunsky
f765e2603b move Torch to optional dependencies (to avoid clash with TF over CuDNN) 2024-12-04 15:57:13 +00:00
vahidrezanezhad
871d7bfc5a fixed: machine based reading order cause tuple index out of range error if number of textregion is one. 2024-12-04 16:41:00 +01:00
vahidrezanezhad
6aad006f4c filter textregions without textline 2024-12-02 12:43:57 +01:00
kba
1083d1c7fb gha: try to free disk space 2024-11-25 19:32:48 +01:00
vahidrezanezhad
8014a9e416
Update Makefile 2024-11-22 19:47:06 +01:00
vahidrezanezhad
3000255a24
Update Makefile 2024-11-22 12:40:21 +01:00
vahidrezanezhad
1746920275
Update Makefile 2024-11-21 12:08:29 +01:00
vahidrezanezhad
b622494f34 new table detection model is integrated 2024-11-21 02:16:22 +01:00
vahidrezanezhad
d9f79c3404 fixing IndexError by reading order detection 2024-11-18 10:15:19 +01:00
vahidrezanezhad
5fa8ca46a4 updating requirements 2024-11-14 17:35:00 +01:00
vahidrezanezhad
ce5b611296 tests are passed - new models by the way should be uploaded 2024-11-14 17:18:07 +01:00
vahidrezanezhad
f43c49c508 textlines of drop capitals are connected to corresponding textline if possible otherwise they are inserted in corresponding textregion 2024-11-13 11:53:56 +01:00
vahidrezanezhad
22b0b07a73 drop capital and marginals extraction is updated 2024-11-11 19:01:40 +01:00
Clemens Neudecker
1ae77e61c8
Update requirements.txt 2024-11-11 14:11:36 +01:00
vahidrezanezhad
8409de0e58 sbb_binarization is integrated into eynollah works in framework of ocrd - sbb_binarization in ocrd works for individual images by the way as standalone flowing from directory can be used now. For eynollah in ocrd framework I have added -light version as default parameter. 2024-11-10 19:34:43 +01:00
vahidrezanezhad
0914b5ff8a resolve merge conflict of main branch with machine based reading order branch 2024-11-06 00:34:00 +01:00
vahidrezanezhad
6aee70d0cd Resolve merge conflict of main and machine based reading order branch 2024-11-06 00:10:25 +01:00
vahidrezanezhad
bceeeb56c1
Merge pull request #138 from qurator-spk/extracting_images_only
Extracting images only
2024-11-05 22:10:51 +01:00
vahidrezanezhad
f7e5fb917f resolving merge conflict of machine based reading order and extracting only images branches 2024-11-05 22:09:39 +01:00
vahidrezanezhad
751b0102f7 updating early layout inference for light version 2024-11-05 19:50:18 +01:00
vahidrezanezhad
e796a99c5c updating inference for early layout in the case of documents with number of columns bigger than 2 2024-10-30 15:02:50 +01:00
vahidrezanezhad
438df52287 updating 2024-10-30 00:52:09 +01:00
vahidrezanezhad
90ee2d61dc textline segmentation is masked with drop capitals 2024-10-28 20:56:06 +01:00
vahidrezanezhad
5037e9896d Merge branch 'machine_based_reading_order_integration' of https://github.com/qurator-spk/eynollah into machine_based_reading_order_integration 2024-10-25 19:47:20 +02:00
vahidrezanezhad
82281bd6cf fixing a bug occuring with reading order + Slro option with no patch textline model and thresholding artificial class 2024-10-25 19:42:48 +02:00
vahidrezanezhad
328d33e3dc Temporary commit – textline prediction without patches 2024-10-23 16:55:41 +02:00
vahidrezanezhad
70772d4104 binarization as a standalone command 2024-10-21 23:46:38 +02:00
vahidrezanezhad
f93fa12441 doing more multiprocessing in order to make the process faster 2024-10-18 09:14:42 +02:00
vahidrezanezhad
3ef4eac24c textlines of textregions are extracted in a faster way + early layout for all documents is done with no patches model and on rgb input 2024-10-17 19:12:28 +02:00
vahidrezanezhad
1da4b7f589 updating light version 2024-10-07 10:55:10 +02:00
vahidrezanezhad
543ed4bc38 -light version need -tll to be enabled otherwise the process will be ended. 2024-10-02 14:09:13 +02:00
Clemens Neudecker
51f6ef63f5
Merge pull request #137 from qurator-spk/dockerfile
Add Dockerfile and make docker target
2024-10-01 17:08:22 +02:00
kba
b13759fdcf ci: smoke-test make docker 2024-10-01 15:38:39 +02:00
kba
c487be2a1d dockerfile: use src-layout 2024-10-01 15:38:01 +02:00
kba
7eb1390583 Merge branch 'main' into dockerfile 2024-10-01 15:25:56 +02:00