qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2025-10-27 07:44:12 +01:00
Code Issues Projects Releases Wiki Activity
94 commits 26 branches 21 tags 23 MiB
Commit graph

94 commits

This branch
This branch
All branches
Author SHA1 Message Date
vahidrezanezhad
aeb2ee4e3e scaling, channels shuffling, rgb background and red content added to no patch augmentation 2024-08-21 19:33:23 +02:00
vahidrezanezhad
445c45cb87 updating augmentations 2024-08-21 16:17:59 +02:00
vahidrezanezhad
5e1821a741 augmentation function for red textlines, rgb background and scaling for no patch case 2024-08-21 00:48:30 +02:00
vahidrezanezhad
bf5837bf6e update 2024-08-09 13:20:09 +02:00
vahidrezanezhad
3b90347a94 save only layout output. different from overlayed layout on image 2024-08-09 12:46:18 +02:00
Clemens Neudecker
2d83b8faad add documentation from wiki as markdown file to the codebase 2024-08-08 16:35:06 +02:00
vahidrezanezhad
6fb28d6ce8 erosion rate changed 2024-08-01 14:30:51 +02:00
vahidrezanezhad
381976099f inference updated 2024-07-24 18:00:39 +02:00
vahidrezanezhad
2c822dae4e erosion and dilation parameters are changed & separators are written in label images after artificial label 2024-07-24 16:52:05 +02:00
b-vr103
840d7c2283 increasing margin in the case of pixelwise inference 2024-07-23 11:29:05 +02:00
b-vr103
861f0b1ebd brightness augmentation modified 2024-07-17 18:20:24 +02:00
vahidrezanezhad
453d0fbf92 adding degrading and brightness augmentation to no patches case training 2024-07-17 17:14:20 +02:00
vahidrezanezhad
3bceec9c19 printspace_as_class_in_layout is integrated. Printspace can be defined as a class for layout segmentation 2024-07-16 18:29:27 +02:00
vahidrezanezhad
9260d2962a resolving typo 2024-07-09 03:04:29 +02:00
vahidrezanezhad
fe69b9c4a8 update inference 2024-06-21 23:42:25 +02:00
vahidrezanezhad
b3cd01de37 update reading order machine based 2024-06-21 13:06:26 +02:00
vahidrezanezhad
66022cf771 update config 2024-06-12 17:40:40 +02:00
vahidrezanezhad
22d7359db2 Transformer+CNN structure is added to vision transformer type 2024-06-12 17:39:57 +02:00
vahidrezanezhad
95faf1a4c8 transformer patch size is dynamic now. 2024-06-12 13:26:27 +02:00
vahidrezanezhad
29da23da76 binarization as a separate task of segmentation 2024-06-11 17:48:30 +02:00
vahidrezanezhad
1921e6754f updating train.py nontransformer backend 2024-06-10 22:15:30 +02:00
vahidrezanezhad
cc91e4b12c updating train.py 2024-06-07 16:24:31 +02:00
vahidrezanezhad
4c376289e9 just defined graphic region types can be extracted as label 2024-06-06 18:55:22 +02:00
vahidrezanezhad
0e4dd0b9ef just defined textregion types can be extracted as label 2024-06-06 18:47:30 +02:00
vahidrezanezhad
5a5914e06c just defined textregion types can be extracted as label 2024-06-06 18:45:47 +02:00
vahidrezanezhad
742e3c2aa2 Update README.md 2024-06-06 14:46:06 +02:00
vahidrezanezhad
13ebe71d13 replacement in a list done correctly 2024-06-06 14:38:29 +02:00
vahidrezanezhad
3ef0dbdd42 scaling and cropping of labels and org images 2024-05-30 16:59:50 +02:00
vahidrezanezhad
47a1646451 modifying xml parsing 2024-05-30 12:56:56 +02:00
vahidrezanezhad
09789619a8 min_area size of regions considered for reading order detection passed as an argument for inference 2024-05-29 13:07:06 +02:00
vahidrezanezhad
06ed006193 reading order detection on xml with layout + result will be written in an output directory with the same file name 2024-05-29 11:18:35 +02:00
vahidrezanezhad
4fb45a6711 inference for reading order 2024-05-28 16:48:51 +02:00
vahidrezanezhad
cc7577d2c1 min area size of text region passes as an argument for machine based reading order 2024-05-28 10:14:16 +02:00
vahidrezanezhad
467bbb2884 pass degrading scales for image enhancement as a json file 2024-05-28 10:01:17 +02:00
vahidrezanezhad
ccf520d3c7 adding rest_as_paragraph and rest_as_graphic to elements 2024-05-27 17:23:49 +02:00
vahidrezanezhad
9638098ae7 machine based reading order training is integrated 2024-05-24 16:39:48 +02:00
vahidrezanezhad
d346b317fb machine based reading order training dataset generator is added 2024-05-24 14:42:58 +02:00
vahidrezanezhad
61487bf782 use case printspace is added 2024-05-23 17:36:23 +02:00
vahidrezanezhad
a83d53c27d use cases like textline, word and glyph are added 2024-05-23 17:14:31 +02:00
vahidrezanezhad
348d323c7c missing text types are added 2024-05-23 15:43:31 +02:00
vahidrezanezhad
47c6bf6b97 dynamic layout decorated with artificial class on text elements boundry 2024-05-23 11:14:14 +02:00
vahidrezanezhad
f1c2913c03 page2label with a dynamic layout 2024-05-22 12:38:24 +02:00
vahidrezanezhad
b2085a1d01 update requirements 2024-05-17 09:08:25 +02:00
vahidrezanezhad
faeac997e1 page to label enable textline new concept 2024-05-17 09:10:13 +02:00
vahidrezanezhad
d6a057ba70 adding page xml to label generator 2024-05-16 15:03:23 +02:00
vahidrezanezhad
d277ec4b31 Update utils.py 2024-05-12 08:32:28 +02:00
vahidrezanezhad
241cb907cb Update train.py 
avoid ensembling if no model weights met the threshold f1 score in the case of classification
 2024-05-08 14:47:16 +02:00
vahidrezanezhad
bc2ca71802 modifications 2024-05-07 16:24:12 +02:00
vahidrezanezhad
e1f62c2e98 inference script is added 2024-05-07 13:34:03 +02:00
vahidrezanezhad
c989f7ac61 adding enhancement training 2024-05-06 18:31:48 +02:00