f9d158fafa
Bump version to 0.3.0 ( #461 )
...
* Bump version to 0.3.0
* update changelog
2021-08-25 16:43:31 +08:00
c0728c49b8
[Docs] Refactor docs ( #409 )
2021-08-25 16:41:07 +08:00
0881c2d2a2
[Fix] Add zero division handler in poly utils, remove Polygon3 ( #448 )
...
* Add check to avoid zero div in iou computation
* replace polygon3 with shapely
* remove req of Polygon3
2021-08-25 13:14:58 +08:00
7c1bf45c63
[Fix] fix inconsisent training/eval state after SyncBN->BN ( #453 )
2021-08-25 13:14:03 +08:00
961fbb6ca5
[Feature] Support persistent_workers option in data loader ( #459 )
...
* Support persistent_workers option in data loader
Signed-off-by: lizz <lizz@sensetime.com>
* Update test.py
* yapf bug
Signed-off-by: lizz <lizz@sensetime.com>
2021-08-25 13:12:53 +08:00
ef111b7c55
[Enhancement] Update demo/ ( #454 )
...
* update demo
* rm demo/readme.md
2021-08-25 03:45:47 +00:00
4e6f70d7c4
[Enhancement] Replace lanms-proper with lanms-neo ( #455 )
2021-08-23 15:32:33 +08:00
d683b14283
[Fix] Totaltext_converter: skip invalid annotations ( #438 )
...
* [Fix] Skip invalid annoataions
2021-08-20 11:23:05 +08:00
8f377f2fa2
[Feature] Add Satrn ( #405 )
...
* Add SATRN
* Create satrn_small_academic.py
* Update README.md
* change config name
* Update mmocr/models/textrecog/backbones/shallow_cnn.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update configs/textrecog/satrn/satrn_academic.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update configs/textrecog/satrn/satrn_small.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update shallow_cnn.py
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update test_ocr_encoder.py
* change keep_aspect_ratio=False
* Update transformer_layer.py
* Update configs/textrecog/satrn/satrn_small.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update configs/textrecog/satrn/satrn_academic.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update mmocr/models/textrecog/layers/transformer_layer.py
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update transformer_layer.py
* Apply suggestions from code review
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
* Update transformer_layer.py
* update satrn readme
* add satrn to ocr.py
* add satrn_sm and fix configs
* add a test for config
* add copyright info
* use mmocr registry
Co-authored-by: Tong Gao <gaotongxiao@gmail.com>
2021-08-19 22:02:58 +08:00
bc4c781e8f
[fix] fix when dbnet polygon points number less than 4 there is a uclip bug ( #443 )
2021-08-19 12:05:28 +00:00
7571763376
[Refactor] Use MMOCR's registry ( #436 )
...
* [Refactor] Use MMOCR's registry
1. Define MMOCR's registries as a child of MMDet's
2. Register all models to MMOCR's own registries
3. Modify some model configs so that some models in MMDet can be
correctly located
4. Remove some outdated demo scripts
* add detectors
2021-08-19 19:17:15 +08:00
b53098dc5c
update citation information ( #440 )
2021-08-19 09:05:30 +00:00
b8f7ead74c
[Enhancement] Add copyright info ( #439 )
...
* add copyright info
2021-08-17 17:39:30 +08:00
9b5b25ef71
[Fix] fix a bug of DBNet when text repr type is poly ( #421 )
...
* fix dbnet bug when text repr type is poly
* add db_decode unit test
2021-08-13 21:39:53 +08:00
80a0536c7c
[Enhancement] Add tests for ocr.py ( #428 )
...
* minor fix to ocr.py
* add test for ocr.py
2021-08-13 09:55:33 +08:00
906faec372
update digit_version and add a test ( #427 )
2021-08-12 08:53:45 +00:00
7bbb14f0d1
[Fix] Replace SyncBN with BN for inference ( #420 )
...
* add revert_sync_batchnorm
* replace SyncBN in inference and test scripts
* add tests
* hide BatchNormXd
2021-08-10 22:19:17 +08:00
532e8f808d
Fix #416 : Fix undesired assignment to "pretrained" in test.py ( #418 )
...
* fix ocr linting
* fix undesired assignment to pretrained key
2021-08-10 14:07:41 +00:00
f62b4513f5
[Fix] Fix a wrong link in ocr.py ( #417 )
2021-08-09 19:52:30 +08:00
304b48e187
[Docs] Full screen support in readthedocs ( #413 )
...
* add full screen css
* New dir structure
2021-08-09 11:16:43 +00:00
8227fec826
[Fix] Fix single model inference in ocr.py ( #411 )
2021-08-04 16:58:47 +08:00
2cbb1a6cd6
[Docs] Fix docs ( #395 )
...
1. Supplement links for ct80 and svtp
2. Simplify getting_started.md
3. clarify batch_mode support in demo.md
2021-08-04 14:26:16 +08:00
7b16754006
[Docs] Fix API reference & version, support exporting pdf & epub ( #406 )
...
* add mmdet
* remove uncessary autodoc mock import
* update docs version
* bind docs version to version.py
* support pdf and epub
* fix readthedocs
2021-08-04 14:24:52 +08:00
b812b7a57c
[Feature] Add PyTorch 1.9 tests in CI ( #401 )
2021-08-04 14:23:53 +08:00
72a5c31577
add mim support ( #394 )
2021-08-04 14:21:41 +08:00
68ec3f5519
Add kie image demo and docs. ( #374 )
...
* Add kie_image_demo.py
* Add kie demo docs
* Add brief instructions for kie image demo
* Add ann file field and return data for kie image demo
* Follow lint and import rules
* Fix bugs, reuse functions in KIEDataset, and use a new demo pic
* Add config-dir and fix indexing bug in ocr script
* [Feature] Improve ocr.py
1. Add box stitching back to ocr.py
2. Add config_dir which allows users to specify the default config path
3. Warn users when overriding parameters are set
4. Allow users to use customized checkpoint files
* Add docs for new ocr.py
* Add docs for merge
* Support kie in ocr.py
* Merged kie to ocr.py
* update docs, remove unsupported unvisual sdmgr
* Update mmocr/apis/inference.py
Co-authored-by: Hongbin Sun <hongbin306@gmail.com>
* Apply suggestions from code review
Co-authored-by: Hongbin Sun <hongbin306@gmail.com>
* fix linting
Co-authored-by: gaotongxiao <gaotongxiao@gmail.com>
Co-authored-by: Hongbin Sun <hongbin306@gmail.com>
2021-08-04 14:20:13 +08:00
f24be6c614
fix bug of dice loss: loss always 1 if empty target ( #408 )
2021-08-04 06:05:31 +00:00
3707d67106
support batch inference for crnn and segocr ( #407 )
...
* support batch inference for crnn and segocr
2021-08-03 16:52:58 +08:00
9c76f7b5cc
[Feature] Improve ocr.py ( #400 )
...
* [Feature] Improve ocr.py
1. Add box stitching back to ocr.py
2. Add config_dir which allows users to specify the default config path
3. Warn users when overriding parameters are set
4. Allow users to use customized checkpoint files
5. Add docs
2021-08-03 15:57:30 +08:00
a2af2b4d31
Add output option to det_recog_show_result method ( #386 )
...
* add out_file arg
* fix small bug
* Update visualize.py
2021-07-25 12:29:40 +08:00
76785e185a
User friendly API v2 + Docs ! ( #371 )
...
* major update
- Refactor code
- Support for folder and list/tuple or np.arrays or img paths
- Better export method
* feature update
- Batch size support
- More refactoring
* added docs
* Optimize docs structure, fix improper layout in readthedocs
Co-authored-by: gaotongxiao <gaotongxiao@gmail.com>
2021-07-25 00:22:27 +08:00
200dfe5fe2
[Fix]: Fix dim mismatch in batch test/inference of DBNet ( #383 )
...
Modification
1.change the dbnet output dim
2.change the input dim for DBloss
2021-07-24 09:14:03 +08:00
02d657d141
[Fix] upgrade version requirement of mmdet to 2.14.0 to avoid known bugs ( #382 )
2021-07-21 18:01:55 +08:00
969c65165f
[Fix] Remove depreciated pretrained in kie_test_imgs ( #381 )
2021-07-21 15:18:59 +08:00
c1ae3a41e5
Bump version to 0.2.1 ( #378 )
2021-07-20 23:19:07 +08:00
d8cb48ab46
[Docs] update dependency version req, dockerfile and change logs for 0.2.1 ( #331 )
...
* update pytorch req and dockerfile
* Update dependency requirement
* update readme for 0.2.1
* update change log
* update release date
2021-07-20 23:18:47 +08:00
4f7270e574
Fix #282 : Support init_cfg & update depreciated configs ( #365 )
...
* update coco ref
* init_cfg for dbnet
* initcfg for mask_rcnn
* textsnake init_cfg
* fix dbnet
* panet initcfg
* psenet initcfg
* fcenet initcfg
* drrg initcfg
* add init_cfg to detectors
* update maskrcnn config file to support mmdet
* fix init_cfg of fce_head
* crnn initcfg
* init_weights in training
* nrtr initcfg
* robust_scanner initcfg
* sar init_cfg
* seg init_cfg
* tps_crnn init_cfg
* sdmgr initcfg
* ner init_cfg
* fix textsnake
* sdmgr initcfg
* move "pretrained" to "init_cfg" for config files
* Moduleslist update
* fix seg
* ner init_cfg
* fix base
* fix encode decode recognizer
* revert dbnet config
* fix crnn
* fix base.py
* fix robust_scanner
* fix panet
* fix test
* remove redundant init_weights() in fcehead
* clean up
* relex mmdet version in workflow
* Add dependency version check
* Update mmocr/models/textdet/dense_heads/pse_head.py
Co-authored-by: Hongbin Sun <hongbin306@gmail.com>
Co-authored-by: Hongbin Sun <hongbin306@gmail.com>
2021-07-20 23:18:25 +08:00
884755d05d
Fix #112 : Remove the need of drop_orientation_info in data preprocessing steps ( #375 )
...
* ctw1500 ignore orientation
* restore maskrcnn config
* ignore_orientation support for icdar datasets
* update docs
* ignore orientation for total text
* Add LoadOCRImageFromFile
* Fix typo
* simplify design
* remove LoadOCRImageFromFile
* update chinese docs
2021-07-20 23:02:25 +08:00
6d97d8286e
Update nrtr batch size ( #370 )
2021-07-16 15:06:36 +08:00
0effc39c5f
[Docs] fix dead links, remove invalid TOC in CN docs ( #367 )
2021-07-15 08:28:36 +00:00
87bd295a20
[Feature] Add User friendly API ( #366 )
2021-07-15 14:18:11 +08:00
782e966c26
fix #347 : add docs_zh_CN/datasets.md ( #362 )
...
* fix #347 : add docs_zh_CN/datasets.md
* fix #347 : translation correction
2021-07-13 12:54:47 +08:00
02e3b98684
fix syntext_converter ( #361 )
2021-07-12 02:07:50 +00:00
44fed8590b
fix bug: remove tail whitespace from gt label ( #358 )
...
* fix bug: remove tail whitespace from gt label
* remove useless
* refactor
2021-07-09 13:24:34 +00:00
243f47dc03
add totaltext for recog and det ( #357 )
...
* add totaltext for recog and det
* add setup
* fix doc
* fix based on comments
2021-07-08 21:52:50 +08:00
1a5a880abb
fix nrtr cfg ( #356 )
2021-07-08 13:30:58 +08:00
500b687469
[Docs] Fix missed deployment in en docs and add language switcher to the sidebar ( #355 )
...
* fix deployment and add language switcher
* change way to switch language and add chinese toc
* update zh_CN datasets.md (to be translated)
2021-07-07 16:21:52 +08:00
68df4fbe80
[Feature] Add synthtext converter and update docs ( #351 )
...
* Add synthtext converter and update docs
* minor docs fix
2021-07-07 15:54:29 +08:00
4fcff1f613
[Docs] Add cn docs framework ( #353 )
...
* add CN demo docs
* add deployment.md to docs
* add placeholder CN docs
* Add language switching hint
2021-07-07 14:13:27 +08:00
19aefa1ae1
Added support for ConcatDataset with two workflows ( #348 )
2021-07-06 16:37:37 +08:00
Tong Gao
lizz
LvTianlei
liukuikun
yuexy
Alireza Kashani
Omkar Manjrekar
Jiaqi Duan
Hongbin Sun
SamAyala
Zichen Tian
quincylin1
Antonio Lanza