Commit Graph

973 Commits (37f44372b11f30032b1c591d08a41ff6d3521c63)

Author SHA1 Message Date
Lin Manhui 2d44a71b20
Toward Devkit Consistency (#10150)
* Accommodate UAPI

* Fix signal handler

* Save model.pdopt

* Change variable name

* Update vdl dir
2023-06-14 14:31:45 +08:00
gmm cdf6d1fb42
fix profile (#10121) 2023-06-08 18:54:29 +08:00
zxcd 46a6950e7a
add dtype param for arange API. (#10075) 2023-06-02 14:21:17 +08:00
huangjun12 0e9c6630ee
fix det v4 bug in dynamic ratio (#9874)
* fix set bug

* refine 960 to 640

* fix details

* add epoch num

* add rep export

* use db head
2023-05-24 14:59:36 +08:00
xiaoting 096fd27121
fix sr_telescope (#10004) 2023-05-23 16:54:52 +08:00
topduke 2c0664b241
fix bug when multi gpus training(sampler) (#9963)
* fix gris_sample data type bug when use fp16

* fix gris_sample data type bug when use fp16

* fix v4rec batchsize

* fix bug of hang when multi gpus training(sampler)
2023-05-17 19:30:33 +08:00
Double_V abc4be007e
add nrtr dml distill loss (#9968)
* support min_area_rect crop

* add check_install

* fix requirement.txt

* fix check_install

* add lanms-neo for drrg

* fix

* fix doc

* fix

* support set gpu_id when inference

* fix #8855

* fix #8855

* opt slim doc

* fix doc bug

* add v4_rec_distill config

* delete debug

* fix comment

* fix comment

* add dml nrtr distill loss
2023-05-17 16:45:04 +08:00
Double_V 1643f268d3
add V4 rec distill (#9921)
* support min_area_rect crop

* add check_install

* fix requirement.txt

* fix check_install

* add lanms-neo for drrg

* fix

* fix doc

* fix

* support set gpu_id when inference

* fix #8855

* fix #8855

* opt slim doc

* fix doc bug

* add v4_rec_distill config

* delete debug

* fix comment

* fix comment
2023-05-15 20:32:48 +08:00
topduke 425166434c
Fix grid_sample data type bug when use fp16 (#9930)
* fix gris_sample data type bug when use fp16

* fix gris_sample data type bug when use fp16

* fix v4rec batchsize
2023-05-15 17:03:53 +08:00
Double_V 24ff4def48
Pfhead (#9898)
* support min_area_rect crop

* add check_install

* fix requirement.txt

* fix check_install

* add lanms-neo for drrg

* fix

* fix doc

* fix

* support set gpu_id when inference

* fix #8855

* fix #8855

* opt slim doc

* fix doc bug

* rename

* rename
2023-05-15 10:57:30 +08:00
xiaoting 7e0c8aea84
revert eval mode (#9843)
* revert eval mode

* update hgnet config
2023-05-04 12:59:55 +08:00
xiaoting b3066812fc
Multi scale (#9837)
* update for multi scale

* update for multi scale

* update for multi scale

* rm notes
2023-04-28 11:04:01 +08:00
xiaoting 26519a6d17
update PPLCNetV3 name (#9802) 2023-04-23 16:13:01 +08:00
huangjun12 ca8c8200ba
add PP-OCRv4 det code (#9766)
* add ppocrv4 det student and teacher model

* update head and config, refine details

* refine config and head details

* refine config and head details

* refine details

* refine details

* remove application

* refine fpn

* fix bug

* update code

* fix bug

* align lcnet to rec

* align hgnet to rec

* refine make shrink

* remove theseus layer
2023-04-21 18:10:26 +08:00
user3984 7710ee04c6
fix data type error when training with fp16 and DynamicToStatic (#9696)
Co-authored-by: WenmuZhou <572459439@qq.com>
2023-04-21 13:53:09 +08:00
topduke 2a98d40b10
Add v4rec hgnet (#9768)
* v4rec code

* v4rec add nrtrloss

* Add V4rec backbone file

* Add V4Rec config file.

* Fix V4rec reparameters when export_model

* convert lvnetv3

* fix codestyle

* fix infer_rec v4rec

* add v4rec hgnet

* add v4rec hgnet config

* add svtr_hgnet

* fix bugs in infer_rec and hgnet
2023-04-21 12:34:48 +08:00
topduke 43abe2fa4b
V4Rec code pr (#9725)
* v4rec code

* v4rec add nrtrloss

* Add V4rec backbone file

* Add V4Rec config file.

* Fix V4rec reparameters when export_model

* convert lvnetv3

* fix codestyle

* fix infer_rec v4rec
2023-04-19 16:04:23 +08:00
zhoujun 20cb993cb6
add d2t train support of svtr (#9479)
* add d2s train for slanet and v3

* fix bug

* udpate tipc to_static

* update db

* remove_print

* update benchmark_train.sh

* update maybe_download_params

* add d2t train support of svtr
2023-03-20 17:26:19 +08:00
xiaoting 1e8b7eb510
rever 'limit maximum resizing ratio' (#9448) 2023-03-15 18:49:45 +08:00
zhoujun 36ec3a406c
update tipc to_static (#9369)
* add d2s train for slanet and v3

* fix bug

* udpate tipc to_static

* update db

* remove_print

* update benchmark_train.sh

* update maybe_download_params
2023-03-10 19:07:37 +08:00
zhoujun 2e05d54af8
add d2s train for slanet and v3 (#9341)
* add d2s train for slanet and v3

* fix bug
2023-03-09 11:21:34 +08:00
zhoujun 623424fce0
support auto download model from bos (#9349) 2023-03-08 19:21:28 +08:00
zhoujun 1139a6c9ec
add table master to benchmark (#9107)
* Add custom detection and recognition model usage instructions in re

* update

* Add custom detection and recognition model usage instructions in re

* add db net for benchmark

* rename benckmark to PaddleOCR_benchmark

* add addict to req

* rename

* add table master benckmark

* support tablemaster d2s train
2023-02-22 15:25:05 +08:00
topduke bfa05707dd
fix nrtr posrprocess lower in chinese (#9103) 2023-02-20 09:39:15 +08:00
Double_V 4fd68e9bd5
fix east post process of dygraph (#9099)
* support min_area_rect crop

* add check_install

* fix requirement.txt

* fix check_install

* add lanms-neo for drrg

* fix

* fix doc

* fix

* support set gpu_id when inference

* fix #8855

* fix #8855

* opt slim doc

* fix doc bug

* fix east postprocess
2023-02-16 12:33:29 +08:00
topduke 72d91eab11
update Blur to fix training time consumption (#9068)
* fix piecewise boundaries and droppath dtype

* fix tipc piecewise boundaries and droppath dtype

* update Blur to fix training time consumption
2023-02-14 16:04:19 +08:00
xiaoting dd02918bb4
rename PP-OCRv3 algorithm to SVTR_LCNet (#9025)
* rename PP-OCRv3 algorithm to SVTR_LCNet

* rename PP-OCRv3 algorithm to SVTR_LCNet

* update multi-lang config for v3
2023-02-13 11:17:58 +08:00
zhiminzhang0830 30201ef954
add satrn (#8433)
* add satrn

* 修复satrn导出问题

* 规范satrn config文件

* 删除SATRNRecResizeImg

---------

Co-authored-by: zhiminzhang0830 <zhangzhimin04@baidu.com>
2023-02-08 11:25:50 +08:00
topduke 8a45b5e08c
Fix piecewise boundaries and droppath dtype (#8988)
* fix piecewise boundaries and droppath dtype

* fix tipc piecewise boundaries and droppath dtype
2023-02-07 13:57:24 +08:00
Double_V 3629d5a365
support set gpu_id when inference (#8926)
* support min_area_rect crop

* add check_install

* fix requirement.txt

* fix check_install

* add lanms-neo for drrg

* fix

* fix doc

* fix

* support set gpu_id when inference

* fix #8855

* fix #8855
2023-01-30 18:46:41 +08:00
andyj 9059f747e6
support new version of numpy (#8921)
* add finetune en doc & test=document_fix

* fix dead link & test=document_fix

* fix dead link & test=document_fix

* update check img

* fix det res dtype

* update args default type & test=document_fix

* fix numpy version

* support numpy1.24.0
2023-01-30 16:05:14 +08:00
topduke 955c46e0c5
add svtr FAQ and data_aug (#8864) 2023-01-19 22:14:28 +08:00
Double_V 414bf18a20
fix issue 9659 and doc (#8690) 2022-12-23 10:02:45 +08:00
xiaoting ca8de0c69e
fix aster loss for axis (#8675) 2022-12-20 13:58:57 +08:00
Double_V 5cac747656
[bug] fix issues about sort_function and DB Head (#8580)
* support min_area_rect crop

* add check_install

* fix requirement.txt

* fix check_install

* add lanms-neo for drrg

* fix

* fix doc

* fix
2022-12-13 16:26:42 +08:00
zhoujun f45c6ec668
mv Polygon import into func (#8506) 2022-12-01 16:02:47 +08:00
Double_V 58a5c0b35d
[check_install] (#8177)
* support min_area_rect crop

* add check_install

* fix requirement.txt

* fix check_install

* add lanms-neo for drrg

* fix

* fix doc
2022-11-15 14:25:39 +08:00
zhoujun 1af190edc5
Adapt to 0-dimensional tensor (#8165) 2022-10-31 19:10:02 +08:00
andyjpaddle 27f4e8a857 rm visionlan invalid params 2022-10-27 07:51:45 +00:00
topduke d810d4859e
Update rec_nrtr_head.py (#8108)
fix bug when export nrtr model with paddlepaddle develop version
2022-10-26 14:19:20 +08:00
littletomatodonkey b92501faf6
fix pic (#8067) 2022-10-24 15:43:01 +08:00
Lizhi Chen 182c1db291
Merge branch 'dygraph' into dygraph 2022-10-20 10:14:55 +08:00
zhoujun 3907c72a08
Merge pull request #7940 from dorren002/new_branch
add handwritten mathematical expression recognition algorithm CAN
2022-10-19 19:28:59 +08:00
andyj 38290a2853
update visionlan (#7989) 2022-10-19 17:12:12 +08:00
贾晓 9d4d44ce07
Merge pull request #7976 from WenmuZhou/tipc3
only in DRRG import DRRGHead
2022-10-18 16:09:15 +08:00
WenmuZhou 273068527f only in DRRG import DRRGHead 2022-10-18 07:31:07 +00:00
dorren ea557c50ee update config 2022-10-17 19:58:35 +08:00
andyj caf77ddb2f
Merge pull request #7957 from andyjpaddle/fix_vl_dict
fix visionlan default dict
2022-10-17 17:21:31 +08:00
dorren f875556541 update can transform method and add copyright info for new file 2022-10-17 17:18:14 +08:00
chenlizhi 9cef24090c add and update sr model 2022-10-17 16:47:28 +08:00
andyjpaddle 9606bec16a fix visionlan default dict 2022-10-17 07:41:36 +00:00
chenlizhi 0cdfc52507 add sr model Text Telescope 2022-10-17 15:15:37 +08:00
dorren c57effb84f update can data loading method and tipc configs, revert precommit config 2022-10-17 15:04:42 +08:00
dorren f3f473d3f9 update CAN model 2022-10-15 20:27:05 +08:00
zhoujun ce72023ce1
Merge pull request #7832 from zhiminzhang0830/drrg_branch
add text detection algorithm DRRG
2022-10-13 13:19:13 +08:00
Double_V 9df7730eeb
Merge pull request #7840 from LDOUBLEV/dygraph
add polygon params
2022-10-12 10:14:16 +08:00
littletomatodonkey 972e7a1ccd
support reconaug (#7873) 2022-10-11 15:13:38 +08:00
zhiminzhang0830 3ee146fa99 Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into drrg_branch 2022-10-10 15:56:08 +08:00
zhiminzhang0830 67ae525e95 删除backbone里的pretrained_model字段 2022-10-10 14:20:28 +08:00
zhoujun 823a8391f1
Merge pull request #7741 from zhiminzhang0830/rfl_branch
add text recognition algorithm rflearning
2022-10-10 14:06:13 +08:00
zhiminzhang0830 483e503826 通过变量类型判断是否是visual 2022-10-10 12:12:47 +08:00
LDOUBLEV 3628ac1c6c fix bug 2022-10-10 10:12:28 +08:00
littletomatodonkey 60559a7ec1
add ser+re hub serving (#7844)
* add ser+re hub serving

* fix
2022-10-09 13:28:12 +08:00
LDOUBLEV 119196f11d add det_box_type params 2022-10-09 11:28:32 +08:00
zhiminzhang0830 c25eec882a Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into rfl_branch 2022-10-09 10:07:46 +08:00
littletomatodonkey eeef62b3c6
fix PREN export and infer (#7833) 2022-10-08 16:37:12 +08:00
zhiminzhang0830 03802c7f93 Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into drrg_branch 2022-10-08 12:00:17 +08:00
zhiminzhang0830 1f9400dd73 add drrg 2022-10-08 11:56:45 +08:00
zhiminzhang0830 c459b72565 添加RFL CNT分支infer支持 2022-10-08 11:20:36 +08:00
WenmuZhou 3e8c78b8c1 fix tablerec-rare train error 2022-09-30 07:01:43 +00:00
zhiminzhang0830 3f8602c169 Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into rfl_branch 2022-09-29 16:40:10 +08:00
zhiminzhang0830 154f42f1b0 推理时添加softmax 2022-09-28 17:44:13 +08:00
zhiminzhang0830 b580fa0517 删除无用函数 2022-09-28 17:43:00 +08:00
zhiminzhang0830 5a380afb1e 添加引用 2022-09-28 17:42:13 +08:00
WenmuZhou fb2b7f5447 add paddle2onnx support for tablerec-rare 2022-09-28 10:32:09 +08:00
zhiminzhang0830 0002349df3 add text recognition algorithm rflearning 2022-09-27 10:54:31 +08:00
WenmuZhou a4bc9da8b6 fix bug 2022-09-21 17:50:50 +08:00
WenmuZhou 06194524ca add re predict 2022-09-20 22:13:27 +08:00
huangjun12 b5268dc3a0 add centripetal text model 2022-09-15 11:08:16 +00:00
Double_V d022b26903
Merge pull request #7420 from wangjingyeye/dyg_db
update pgnet
2022-09-05 18:45:44 +08:00
wangjingyeye 0fd122b674 update pgnet 2022-09-05 09:06:17 +00:00
WenmuZhou 6419c7bdaa add stop_gradient to create tensor 2022-09-05 15:31:19 +08:00
wangjingyeye 4c0b08733d update pgnet 2022-09-05 07:03:16 +00:00
WenmuZhou f6698a32d8 support onnx infer of SLANet 2022-09-05 15:02:22 +08:00
LDOUBLEV 2189e54ad9 padding when image.h+w < 32 2022-09-02 10:45:33 +08:00
wangjingyeye 929b4f4557 update pgnet 2022-08-30 05:58:39 +00:00
LDOUBLEV f630700a99 fix cml 2022-08-29 14:03:05 +08:00
jinming.teo 4af5f63d44 limit maximum resizing ratio 2022-08-26 22:17:23 +08:00
贾晓 ba04894d31
Merge pull request #7344 from Topdu/dygraph
fix arabic postprocess bug
2022-08-25 14:14:58 +08:00
topduke 4f7b28e89f
Update rec_postprocess.py 2022-08-25 12:52:45 +08:00
Topdu 6cded42b10 fix arabic postprocess bug 2022-08-25 04:33:42 +00:00
MissPenguin e6c6250493
Merge pull request #7208 from Topdu/dygraph
update arabic rec model & add pred reverse function
2022-08-24 16:28:29 +08:00
littletomatodonkey 3c1f5df8e4 fix name 2022-08-24 05:55:06 +00:00
littletomatodonkey dd2f7de431 fix name 2022-08-24 05:53:42 +00:00
WenmuZhou 2b1885f159 Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into tipc 2022-08-22 11:33:51 +00:00
WenmuZhou c4fd53070e update fp16 load 2022-08-22 11:33:00 +00:00
Topdu 1a5e1be1fa arabic reverse function 2022-08-22 09:57:57 +00:00
Topdu 2741a84de7 arabic reverse function 2022-08-22 09:53:04 +00:00
WenmuZhou 66f4ae0261 fix is_nlp_model not define error in save_model 2022-08-22 02:25:24 +00:00
littletomatodonkey 988f6cadac
fix kie doc (#7275)
* fix kie doc

* fix en
2022-08-22 09:52:23 +08:00
Evezerest 7e4e87dd6d
cherry pick PRs from community (#7273)
* Merge pull request #6824 from ChenNima/release/2.5-kie-save-res

[kie]add write_kie_result to kie infer tool

* Merge pull request #6677 from TonyJiangWJ/release/2.5

修复内存泄露问题

* Update native.cpp (#6650)

fix issue 6640

* Merge pull request #6625 from ynjang/ynjang

update sorted_boxes

* fix DeprecationWarning, (#6604)

DeprecationWarning: The binary mode of fromstring is deprecated, as it behaves surprisingly on unicode inputs. Use frombuffer instead

* Merge pull request #6585 from maxbachmann/release/2.5

replace GPL licensed components

* Merge pull request #6575 from Eling486/release/2.5

update win doc

* Merge pull request #6477 from MikoyChinese/fix-copy-paste

Fix copy_paste no texts augment.

* Merge pull request #6361 from mohamadmansourX/patch-9

Update README_en.md

Co-authored-by: Double_V <liuvv0203@163.com>
Co-authored-by: shawn <1021362695@qq.com>
Co-authored-by: paopjian <672034519@qq.com>
2022-08-21 18:03:57 +08:00
littletomatodonkey 8c7c45420f
polish kie doc and code (#7255)
* add fapiao kie

* fix readme

* fix fanli

* add readme

* add how to do kie en

* add algo kie

* add algo overview en

* rename vqa to kie

* fix read gif
2022-08-21 10:55:49 +08:00
user1018 b7d99acd2e
update recovery (#7259)
* update recovery

* update recovery

* update recovery

* update recovery

* update recovery
2022-08-19 20:15:37 +08:00
user1018 bfff01db7e
Upstream/layout update (#7220)
* update layout

* update layout

* update layout

* update layout

* update layout
2022-08-18 19:43:03 +08:00
smilelite 40c45e2ccc
Merge branch 'dygraph' into robustscanner_branch 2022-08-17 21:49:11 +08:00
WenmuZhou 604f531b9a add layout model 2022-08-17 04:40:07 +00:00
WenmuZhou bb53c8d100 add table model link 2022-08-16 10:46:09 +00:00
WenmuZhou 22c13af65b Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into table_pr 2022-08-16 03:35:46 +00:00
WenmuZhou ec22e60cb4 fix boxlabel error 2022-08-16 03:34:40 +00:00
smilelite a8ba8963e0 merge cpnflict 0815 2022-08-15 23:03:21 +08:00
Topdu 09ff249363 update arabic rec model & add pred reverse function 2022-08-15 10:42:02 +00:00
WenmuZhou 4369552ea2 add in and out box_format 2022-08-15 10:04:06 +00:00
andyjpaddle c9504a9425 fix vl predict format 2022-08-15 09:38:56 +00:00
andyjpaddle e5e787018e fix vl 2022-08-15 09:05:32 +00:00
WenmuZhou ebe3e885c0 Support variable length input 2022-08-15 09:00:07 +00:00
littletomatodonkey 78871cfefb
kie doc (#7154)
* kie doc

* fix xlm model export

* fix doc

* add wildreceipt dataset

* fix doc

* fix doc
2022-08-15 11:39:11 +08:00
WenmuZhou ba284a2b74 Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into table_pr 2022-08-14 09:07:19 +00:00
WenmuZhou ce321153e7 rm unused code 2022-08-14 09:01:49 +00:00
xiaoting 7054013004
Submit SR model (#6933)
* add sr model

* update for eval

* submit sr

* polish code

* polish code

* polish code

* update sr model

* update doc

* update doc

* update doc

* fix typo

* format code

* update metric

* fix export
2022-08-12 10:49:54 +08:00
smilelite 60ea689df4
Merge branch 'dygraph' into robustscanner_branch 2022-08-11 21:45:21 +08:00
WenmuZhou 73c77ff79d add image_orientation and update quickstart 2022-08-11 10:56:19 +00:00
WenmuZhou c2c43bb1bc rename SLANetLoss to SLALoss 2022-08-10 14:58:08 +00:00
WenmuZhou 70145c3cdb Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into table_pr 2022-08-10 11:01:21 +00:00
an1018 c94025f266 update layout 2022-08-10 12:54:25 +08:00
WenmuZhou d5ea6f21f8 Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into table_pr 2022-08-09 11:36:49 +00:00
WenmuZhou 92f8552121 add "<td></td>" to dict when "<td></td>" not in file 2022-08-09 10:50:30 +00:00
andyjpaddle 3f65b360ef
Add rec algo VisionLAN (#6943)
* add vl

* add vl

* add vl

* add ref

* fix head out

* add visionlan doc

* fix vl infer

* update dict
2022-08-09 11:29:43 +08:00
andyjpaddle 9e1a77ead1 update dict 2022-08-09 01:47:57 +00:00
andyjpaddle 3380010d9a update dict 2022-08-09 01:44:16 +00:00
andyjpaddle 44852aae0c
Merge pull request #7109 from WenmuZhou/tipc1
convert fp16 params to fp32 when params is fp16 format
2022-08-08 19:30:44 +08:00
WenmuZhou 6fe01d43dc Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into table_pr 2022-08-08 06:58:15 +00:00
WenmuZhou ddaa2c2552 add SLANet 2022-08-08 06:50:27 +00:00
smilelite 47a27071eb
Merge branch 'dygraph' into robustscanner_branch 2022-08-08 09:31:16 +08:00
littletomatodonkey 9e4ae9dc12
add vqa code (#7096)
* add vqa code

* add order ocr info

* rename tb-yx order

* polish configs

* add trt offline-tuning

* fix seed and remove unused configs
2022-08-06 15:41:20 +08:00
WenmuZhou c6738f4c53 convert fp16 params to fp32 when params is fp16 format 2022-08-05 12:50:25 +08:00
WenmuZhou d0efcc74c9 convert fp16 params to fp32 when params is fp16 format 2022-08-05 12:41:03 +08:00
WenmuZhou 160e5e384a rename 2.0 to 2_0 2022-08-02 19:41:29 +08:00
andyjpaddle ab0b4e7474 fix vl infer 2022-08-02 11:39:51 +00:00
smilelite 3337dee65a fixed rec_img_aug 20220801 2022-08-01 22:09:12 +08:00
andyjpaddle 7a706ada64 Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into dygraph 2022-07-28 09:14:21 +00:00
andyjpaddle 0d3c2924b1 fix head out 2022-07-28 08:37:11 +00:00
smilelite b781838015 fix conflicts 2022-07-27 23:02:24 +08:00
zhoujun 1696b36bdb
Merge pull request #6841 from smilelite/spin_branch
添加SPIN(第二次修改)
2022-07-26 20:09:34 +08:00
贾晓 db5d66f9de
Merge pull request #6811 from Topdu/dygraph
Fix windows bug of abinet
2022-07-20 19:55:50 +08:00
andyjpaddle 8656a1dd0c add ref 2022-07-20 11:18:01 +00:00
Topdu bb7a3b0e9a fix win&mac bug of abinet | validation 2022-07-20 08:57:55 +00:00
andyjpaddle a3a095150e add vl 2022-07-20 06:21:16 +00:00
andyjpaddle 0401e5203e add vl 2022-07-19 12:54:17 +00:00
andyjpaddle cf533b65c5 add vl 2022-07-19 12:38:54 +00:00
smilelite 7c5b47069e modified config and resnet31 2022-07-14 23:11:20 +08:00