m53297601
/
PaddleOCR
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

dygraph first commit

Browse Source
release/2.0-rc1-0
WenmuZhou 4 years ago
parent 10f7e5192d
commit aad3093a91
148 changed files with 5638 additions and 26873 deletions
Show Stats Download Patch File Download Diff File

22
configs/det/det_db_icdar15_reader.yml
Unescape View File

@ -1,22 +0,0 @@
TrainReader:
reader_function: ppocr.data.det.dataset_traversal,TrainReader
process_function: ppocr.data.det.db_process,DBProcessTrain
num_workers: 8
img_set_dir: ./train_data/icdar2015/text_localization/
label_file_path: ./train_data/icdar2015/text_localization/train_icdar2015_label.txt
EvalReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.db_process,DBProcessTest
img_set_dir: ./train_data/icdar2015/text_localization/
label_file_path: ./train_data/icdar2015/text_localization/test_icdar2015_label.txt
test_image_shape: [736, 1280]
TestReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.db_process,DBProcessTest
infer_img:
img_set_dir: ./train_data/icdar2015/text_localization/
label_file_path: ./train_data/icdar2015/text_localization/test_icdar2015_label.txt
test_image_shape: [736, 1280]
do_eval: True

23
configs/det/det_east_icdar15_reader.yml
Unescape View File

@ -1,23 +0,0 @@
TrainReader:
reader_function: ppocr.data.det.dataset_traversal,TrainReader
process_function: ppocr.data.det.east_process,EASTProcessTrain
num_workers: 8
img_set_dir: ./train_data/icdar2015/text_localization/
label_file_path: ./train_data/icdar2015/text_localization/train_icdar2015_label.txt
background_ratio: 0.125
min_crop_side_ratio: 0.1
min_text_size: 10
EvalReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.east_process,EASTProcessTest
img_set_dir: ./train_data/icdar2015/text_localization/
label_file_path: ./train_data/icdar2015/text_localization/test_icdar2015_label.txt
TestReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.east_process,EASTProcessTest
infer_img:
img_set_dir: ./train_data/icdar2015/text_localization/
label_file_path: ./train_data/icdar2015/text_localization/test_icdar2015_label.txt
do_eval: True

141
configs/det/det_mv3_db.yml
Unescape View File

@ -1,54 +1,133 @@
Global: Global:
algorithm: DB
use_gpu: true use_gpu: true
epoch_num: 1200 epoch_num: 1200
log_smooth_window: 20 log_smooth_window: 20
print_batch_step: 2 print_batch_step: 2
save_model_dir: ./output/det_db/ save_model_dir: ./output/20201010/
save_epoch_step: 200 save_epoch_step: 1200
# evaluation is run every 5000 iterations after the 4000th iteration # evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: [4000, 5000] eval_batch_step: 8
train_batch_size_per_card: 16 # if pretrained_model is saved in static mode, load_static_weights must set to True
test_batch_size_per_card: 16 load_static_weights: True
image_shape: [3, 640, 640] cal_metric_during_train: False
reader_yml: ./configs/det/det_db_icdar15_reader.yml pretrained_model: /home/zhoujun20/pretrain_models/MobileNetV3_large_x0_5_pretrained
pretrain_weights: ./pretrain_models/MobileNetV3_large_x0_5_pretrained/ checkpoints: #./output/det_db_0.001_DiceLoss_256_pp_config_2.0b_4gpu/best_accuracy
checkpoints:
save_res_path: ./output/det_db/predicts_db.txt
save_inference_dir: save_inference_dir:
use_visualdl: True
infer_img: doc/imgs_en/img_10.jpg
save_res_path: ./output/det_db/predicts_db.txt
Architecture: Optimizer:
function: ppocr.modeling.architectures.det_model,DetModel name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
# name: Cosine
lr: 0.001
# warmup_epoch: 0
regularizer:
name: 'L2'
factor: 0
Backbone: Architecture:
function: ppocr.modeling.backbones.det_mobilenet_v3,MobileNetV3 type: det
algorithm: DB
Transform:
Backbone:
name: MobileNetV3
scale: 0.5 scale: 0.5
model_name: large model_name: large
Neck:
Head: name: FPN
function: ppocr.modeling.heads.det_db_head,DBHead out_channels: 256
model_name: large Head:
name: DBHead
k: 50 k: 50
inner_channels: 96
out_channels: 2
Loss: Loss:
function: ppocr.modeling.losses.det_db_loss,DBLoss name: DBLoss
balance_loss: true balance_loss: true
main_loss_type: DiceLoss main_loss_type: DiceLoss
alpha: 5 alpha: 5
beta: 10 beta: 10
ohem_ratio: 3 ohem_ratio: 3
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999
PostProcess: PostProcess:
function: ppocr.postprocess.db_postprocess,DBPostProcess name: DBPostProcess
thresh: 0.3 thresh: 0.3
box_thresh: 0.7 box_thresh: 0.6
max_candidates: 1000 max_candidates: 1000
unclip_ratio: 2.0 unclip_ratio: 1.5
Metric:
name: DetMetric
main_indicator: hmean
TRAIN:
dataset:
name: SimpleDataSet
data_dir: /home/zhoujun20/detection/
file_list:
- /home/zhoujun20/detection/train_icdar2015_label.txt # dataset1
ratio_list: [1.0]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- DetLabelEncode: # Class handling label
- IaaAugment:
augmenter_args:
- { 'type': Fliplr, 'args': { 'p': 0.5 } }
- { 'type': Affine, 'args': { 'rotate': [ -10,10 ] } }
- { 'type': Resize,'args': { 'size': [ 0.5,3 ] } }
- EastRandomCropData:
size: [ 640,640 ]
max_tries: 50
keep_ratio: true
- MakeBorderMap:
shrink_ratio: 0.4
thresh_min: 0.3
thresh_max: 0.7
- MakeShrinkMap:
shrink_ratio: 0.4
min_text_size: 8
- NormalizeImage:
scale: 1./255.
mean: [ 0.485, 0.456, 0.406 ]
std: [ 0.229, 0.224, 0.225 ]
order: 'hwc'
- ToCHWImage:
- keepKeys:
keep_keys: ['image','threshold_map','threshold_mask','shrink_map','shrink_mask'] # dataloader将按照此顺序返回list
loader:
shuffle: True
drop_last: False
batch_size: 16
num_workers: 6
EVAL:
dataset:
name: SimpleDataSet
data_dir: /home/zhoujun20/detection/
file_list:
- /home/zhoujun20/detection/test_icdar2015_label.txt
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- DetLabelEncode: # Class handling label
- DetResizeForTest:
image_shape: [736,1280]
- NormalizeImage:
scale: 1./255.
mean: [ 0.485, 0.456, 0.406 ]
std: [ 0.229, 0.224, 0.225 ]
order: 'hwc'
- ToCHWImage:
- keepKeys:
keep_keys: ['image','shape','polys','ignore_tags']
loader:
shuffle: False
drop_last: False
batch_size: 1 # must be 1
num_workers: 6

45
configs/det/det_mv3_east.yml
Unescape View File

@ -1,45 +0,0 @@
Global:
algorithm: EAST
use_gpu: true
epoch_num: 100000
log_smooth_window: 20
print_batch_step: 5
save_model_dir: ./output/det_east/
save_epoch_step: 200
eval_batch_step: [5000, 5000]
train_batch_size_per_card: 16
test_batch_size_per_card: 16
image_shape: [3, 512, 512]
reader_yml: ./configs/det/det_east_icdar15_reader.yml
pretrain_weights: ./pretrain_models/MobileNetV3_large_x0_5_pretrained/
checkpoints:
save_res_path: ./output/det_east/predicts_east.txt
save_inference_dir:
Architecture:
function: ppocr.modeling.architectures.det_model,DetModel
Backbone:
function: ppocr.modeling.backbones.det_mobilenet_v3,MobileNetV3
scale: 0.5
model_name: large
Head:
function: ppocr.modeling.heads.det_east_head,EASTHead
model_name: small
Loss:
function: ppocr.modeling.losses.det_east_loss,EASTLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999
PostProcess:
function: ppocr.postprocess.east_postprocess,EASTPostPocess
score_thresh: 0.8
cover_thresh: 0.1
nms_thresh: 0.2

139
configs/det/det_r50_vd_db.yml
Unescape View File

@ -1,53 +1,132 @@
Global: Global:
algorithm: DB
use_gpu: true use_gpu: true
epoch_num: 1200 epoch_num: 1200
log_smooth_window: 20 log_smooth_window: 20
print_batch_step: 2 print_batch_step: 2
save_model_dir: ./output/det_db/ save_model_dir: ./output/20201010/
save_epoch_step: 200 save_epoch_step: 1200
eval_batch_step: [5000, 5000] # evaluation is run every 5000 iterations after the 4000th iteration
train_batch_size_per_card: 8 eval_batch_step: 8
test_batch_size_per_card: 16 # if pretrained_model is saved in static mode, load_static_weights must set to True
image_shape: [3, 640, 640] load_static_weights: True
reader_yml: ./configs/det/det_db_icdar15_reader.yml cal_metric_during_train: False
pretrain_weights: ./pretrain_models/ResNet50_vd_ssld_pretrained/ pretrained_model: /home/zhoujun20/pretrain_models/MobileNetV3_large_x0_5_pretrained
save_res_path: ./output/det_db/predicts_db.txt checkpoints: #./output/det_db_0.001_DiceLoss_256_pp_config_2.0b_4gpu/best_accuracy
checkpoints:
save_inference_dir: save_inference_dir:
use_visualdl: True
infer_img: doc/imgs_en/img_10.jpg
save_res_path: ./output/det_db/predicts_db.txt
Architecture: Optimizer:
function: ppocr.modeling.architectures.det_model,DetModel name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
# name: Cosine
lr: 0.001
# warmup_epoch: 0
regularizer:
name: 'L2'
factor: 0
Backbone: Architecture:
function: ppocr.modeling.backbones.det_resnet_vd,ResNet type: det
algorithm: DB
Transform:
Backbone:
name: ResNet
layers: 50 layers: 50
Neck:
Head: name: FPN
function: ppocr.modeling.heads.det_db_head,DBHead out_channels: 256
model_name: large Head:
name: DBHead
k: 50 k: 50
inner_channels: 256
out_channels: 2
Loss: Loss:
function: ppocr.modeling.losses.det_db_loss,DBLoss name: DBLoss
balance_loss: true balance_loss: true
main_loss_type: DiceLoss main_loss_type: DiceLoss
alpha: 5 alpha: 5
beta: 10 beta: 10
ohem_ratio: 3 ohem_ratio: 3
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999
PostProcess: PostProcess:
function: ppocr.postprocess.db_postprocess,DBPostProcess name: DBPostProcess
thresh: 0.3 thresh: 0.3
box_thresh: 0.7 box_thresh: 0.6
max_candidates: 1000 max_candidates: 1000
unclip_ratio: 1.5 unclip_ratio: 1.5
Metric:
name: DetMetric
main_indicator: hmean
TRAIN:
dataset:
name: SimpleDataSet
data_dir: /home/zhoujun20/detection/
file_list:
- /home/zhoujun20/detection/train_icdar2015_label.txt # dataset1
ratio_list: [1.0]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- DetLabelEncode: # Class handling label
- IaaAugment:
augmenter_args:
- { 'type': Fliplr, 'args': { 'p': 0.5 } }
- { 'type': Affine, 'args': { 'rotate': [ -10,10 ] } }
- { 'type': Resize,'args': { 'size': [ 0.5,3 ] } }
- EastRandomCropData:
size: [ 640,640 ]
max_tries: 50
keep_ratio: true
- MakeBorderMap:
shrink_ratio: 0.4
thresh_min: 0.3
thresh_max: 0.7
- MakeShrinkMap:
shrink_ratio: 0.4
min_text_size: 8
- NormalizeImage:
scale: 1./255.
mean: [ 0.485, 0.456, 0.406 ]
std: [ 0.229, 0.224, 0.225 ]
order: 'hwc'
- ToCHWImage:
- keepKeys:
keep_keys: ['image','threshold_map','threshold_mask','shrink_map','shrink_mask'] # dataloader将按照此顺序返回list
loader:
shuffle: True
drop_last: False
batch_size: 16
num_workers: 6
EVAL:
dataset:
name: SimpleDataSet
data_dir: /home/zhoujun20/detection/
file_list:
- /home/zhoujun20/detection/test_icdar2015_label.txt
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- DetLabelEncode: # Class handling label
- DetResizeForTest:
image_shape: [736,1280]
- NormalizeImage:
scale: 1./255.
mean: [ 0.485, 0.456, 0.406 ]
std: [ 0.229, 0.224, 0.225 ]
order: 'hwc'
- ToCHWImage:
- keepKeys:
keep_keys: ['image','shape','polys','ignore_tags']
loader:
shuffle: False
drop_last: False
batch_size: 1 # must be 1
num_workers: 6

44
configs/det/det_r50_vd_east.yml
Unescape View File

@ -1,44 +0,0 @@
Global:
algorithm: EAST
use_gpu: true
epoch_num: 100000
log_smooth_window: 20
print_batch_step: 5
save_model_dir: ./output/det_east/
save_epoch_step: 200
eval_batch_step: [5000, 5000]
train_batch_size_per_card: 8
test_batch_size_per_card: 16
image_shape: [3, 512, 512]
reader_yml: ./configs/det/det_east_icdar15_reader.yml
pretrain_weights: ./pretrain_models/ResNet50_vd_ssld_pretrained/
save_res_path: ./output/det_east/predicts_east.txt
checkpoints:
save_inference_dir:
Architecture:
function: ppocr.modeling.architectures.det_model,DetModel
Backbone:
function: ppocr.modeling.backbones.det_resnet_vd,ResNet
layers: 50
Head:
function: ppocr.modeling.heads.det_east_head,EASTHead
model_name: large
Loss:
function: ppocr.modeling.losses.det_east_loss,EASTLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999
PostProcess:
function: ppocr.postprocess.east_postprocess,EASTPostPocess
score_thresh: 0.8
cover_thresh: 0.1
nms_thresh: 0.2

50
configs/det/det_r50_vd_sast_icdar15.yml
Unescape View File

@ -1,50 +0,0 @@
Global:
algorithm: SAST
use_gpu: true
epoch_num: 2000
log_smooth_window: 20
print_batch_step: 2
save_model_dir: ./output/det_sast/
save_epoch_step: 20
eval_batch_step: 5000
train_batch_size_per_card: 8
test_batch_size_per_card: 8
image_shape: [3, 512, 512]
reader_yml: ./configs/det/det_sast_icdar15_reader.yml
pretrain_weights: ./pretrain_models/ResNet50_vd_ssld_pretrained/
save_res_path: ./output/det_sast/predicts_sast.txt
checkpoints:
save_inference_dir:
Architecture:
function: ppocr.modeling.architectures.det_model,DetModel
Backbone:
function: ppocr.modeling.backbones.det_resnet_vd_sast,ResNet
layers: 50
Head:
function: ppocr.modeling.heads.det_sast_head,SASTHead
model_name: large
only_fpn_up: False
# with_cab: False
with_cab: True
Loss:
function: ppocr.modeling.losses.det_sast_loss,SASTLoss
Optimizer:
function: ppocr.optimizer,RMSProp
base_lr: 0.001
decay:
function: piecewise_decay
boundaries: [30000, 50000, 80000, 100000, 150000]
decay_rate: 0.3
PostProcess:
function: ppocr.postprocess.sast_postprocess,SASTPostProcess
score_thresh: 0.5
sample_pts_num: 2
nms_thresh: 0.2
expand_scale: 1.0
shrink_ratio_of_width: 0.3

50
configs/det/det_r50_vd_sast_totaltext.yml
Unescape View File

@ -1,50 +0,0 @@
Global:
algorithm: SAST
use_gpu: true
epoch_num: 2000
log_smooth_window: 20
print_batch_step: 2
save_model_dir: ./output/det_sast/
save_epoch_step: 20
eval_batch_step: 5000
train_batch_size_per_card: 8
test_batch_size_per_card: 1
image_shape: [3, 512, 512]
reader_yml: ./configs/det/det_sast_totaltext_reader.yml
pretrain_weights: ./pretrain_models/ResNet50_vd_ssld_pretrained/
save_res_path: ./output/det_sast/predicts_sast.txt
checkpoints:
save_inference_dir:
Architecture:
function: ppocr.modeling.architectures.det_model,DetModel
Backbone:
function: ppocr.modeling.backbones.det_resnet_vd_sast,ResNet
layers: 50
Head:
function: ppocr.modeling.heads.det_sast_head,SASTHead
model_name: large
only_fpn_up: False
# with_cab: False
with_cab: True
Loss:
function: ppocr.modeling.losses.det_sast_loss,SASTLoss
Optimizer:
function: ppocr.optimizer,RMSProp
base_lr: 0.001
decay:
function: piecewise_decay
boundaries: [30000, 50000, 80000, 100000, 150000]
decay_rate: 0.3
PostProcess:
function: ppocr.postprocess.sast_postprocess,SASTPostProcess
score_thresh: 0.5
sample_pts_num: 6
nms_thresh: 0.2
expand_scale: 1.2
shrink_ratio_of_width: 0.2

24
configs/det/det_sast_icdar15_reader.yml
Unescape View File

@ -1,24 +0,0 @@
TrainReader:
reader_function: ppocr.data.det.dataset_traversal,TrainReader
process_function: ppocr.data.det.sast_process,SASTProcessTrain
num_workers: 8
img_set_dir: ./train_data/
label_file_path: [./train_data/icdar2013/train_label_json.txt, ./train_data/icdar2015/train_label_json.txt, ./train_data/icdar17_mlt_latin/train_label_json.txt, ./train_data/coco_text_icdar_4pts/train_label_json.txt]
data_ratio_list: [0.1, 0.45, 0.3, 0.15]
min_crop_side_ratio: 0.3
min_crop_size: 24
min_text_size: 4
max_text_size: 512
EvalReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.sast_process,SASTProcessTest
img_set_dir: ./train_data/icdar2015/text_localization/
label_file_path: ./train_data/icdar2015/text_localization/test_icdar2015_label.txt
max_side_len: 1536
TestReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.sast_process,SASTProcessTest
infer_img: ./train_data/icdar2015/text_localization/ch4_test_images/img_11.jpg
max_side_len: 1536

24
configs/det/det_sast_totaltext_reader.yml
Unescape View File

@ -1,24 +0,0 @@
TrainReader:
reader_function: ppocr.data.det.dataset_traversal,TrainReader
process_function: ppocr.data.det.sast_process,SASTProcessTrain
num_workers: 8
img_set_dir: ./train_data/
label_file_path: [./train_data/art_latin_icdar_14pt/train_no_tt_test/train_label_json.txt, ./train_data/total_text_icdar_14pt/train_label_json.txt]
data_ratio_list: [0.5, 0.5]
min_crop_side_ratio: 0.3
min_crop_size: 24
min_text_size: 4
max_text_size: 512
EvalReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.sast_process,SASTProcessTest
img_set_dir: ./train_data/
label_file_path: ./train_data/total_text_icdar_14pt/test_label_json.txt
max_side_len: 768
TestReader:
reader_function: ppocr.data.det.dataset_traversal,EvalTestReader
process_function: ppocr.data.det.sast_process,SASTProcessTest
infer_img: ./train_data/afs/total_text/Images/Test/img623.jpg
max_side_len: 768

12
configs/rec/rec_benchmark_reader.yml
Unescape View File

@ -1,12 +0,0 @@
TrainReader:
reader_function: ppocr.data.rec.dataset_traversal,LMDBReader
num_workers: 8
lmdb_sets_dir: ./train_data/data_lmdb_release/training/
EvalReader:
reader_function: ppocr.data.rec.dataset_traversal,LMDBReader
lmdb_sets_dir: ./train_data/data_lmdb_release/validation/
TestReader:
reader_function: ppocr.data.rec.dataset_traversal,LMDBReader
lmdb_sets_dir: ./train_data/data_lmdb_release/evaluation/

45
configs/rec/rec_chinese_common_train.yml
Unescape View File

@ -1,45 +0,0 @@
Global:
algorithm: CRNN
use_gpu: true
epoch_num: 3000
log_smooth_window: 20
print_batch_step: 10
save_model_dir: ./output/rec_CRNN
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 128
test_batch_size_per_card: 128
image_shape: [3, 32, 320]
max_text_length: 25
character_type: ch
character_dict_path: ./ppocr/utils/ppocr_keys_v1.txt
loss_type: ctc
distort: false
use_space_char: false
reader_yml: ./configs/rec/rec_chinese_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
Backbone:
function: ppocr.modeling.backbones.rec_resnet_vd,ResNet
layers: 34
Head:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
encoder_type: rnn
SeqRNN:
hidden_size: 256
Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.0005
beta1: 0.9
beta2: 0.999

46
configs/rec/rec_chinese_lite_train.yml
Unescape View File

@ -1,46 +0,0 @@
Global:
algorithm: CRNN
use_gpu: true
epoch_num: 3000
log_smooth_window: 20
print_batch_step: 10
save_model_dir: ./output/rec_CRNN
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 320]
max_text_length: 25
character_type: ch
character_dict_path: ./ppocr/utils/ppocr_keys_v1.txt
loss_type: ctc
distort: false
use_space_char: false
reader_yml: ./configs/rec/rec_chinese_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
Backbone:
function: ppocr.modeling.backbones.rec_mobilenet_v3,MobileNetV3
scale: 0.5
model_name: small
Head:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
encoder_type: rnn
SeqRNN:
hidden_size: 48
Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.0005
beta1: 0.9
beta2: 0.999

13
configs/rec/rec_chinese_reader.yml
Unescape View File

@ -1,13 +0,0 @@
TrainReader:
reader_function: ppocr.data.rec.dataset_traversal,SimpleReader
num_workers: 8
img_set_dir: ./train_data
label_file_path: ./train_data/rec_gt_train.txt
EvalReader:
reader_function: ppocr.data.rec.dataset_traversal,SimpleReader
img_set_dir: ./train_data
label_file_path: ./train_data/rec_gt_test.txt
TestReader:
reader_function: ppocr.data.rec.dataset_traversal,SimpleReader

13
configs/rec/rec_icdar15_reader.yml
Unescape View File

@ -1,13 +0,0 @@
TrainReader:
reader_function: ppocr.data.rec.dataset_traversal,SimpleReader
num_workers: 8
img_set_dir: ./train_data/ic15_data
label_file_path: ./train_data/ic15_data/rec_gt_train.txt
EvalReader:
reader_function: ppocr.data.rec.dataset_traversal,SimpleReader
img_set_dir: ./train_data/ic15_data
label_file_path: ./train_data/ic15_data/rec_gt_test.txt
TestReader:
reader_function: ppocr.data.rec.dataset_traversal,SimpleReader

49
configs/rec/rec_icdar15_train.yml
Unescape View File

@ -1,49 +0,0 @@
Global:
algorithm: CRNN
use_gpu: true
epoch_num: 1000
log_smooth_window: 20
print_batch_step: 10
save_model_dir: ./output/rec_CRNN
save_epoch_step: 300
eval_batch_step: 500
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 100]
max_text_length: 25
character_type: en
loss_type: ctc
distort: true
debug: false
reader_yml: ./configs/rec/rec_icdar15_reader.yml
pretrain_weights: ./pretrain_models/rec_mv3_none_bilstm_ctc/best_accuracy
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
Backbone:
function: ppocr.modeling.backbones.rec_mobilenet_v3,MobileNetV3
scale: 0.5
model_name: large
Head:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
encoder_type: rnn
SeqRNN:
hidden_size: 96
Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.0005
beta1: 0.9
beta2: 0.999
decay:
function: cosine_decay
step_each_epoch: 20
total_epoch: 1000

129
configs/rec/rec_mv3_none_bilstm_ctc.yml
Unescape View File

@ -1,43 +1,108 @@
Global: Global:
algorithm: CRNN use_gpu: false
use_gpu: true epoch_num: 500
epoch_num: 72
log_smooth_window: 20 log_smooth_window: 20
print_batch_step: 10 print_batch_step: 10
save_model_dir: output/rec_CRNN save_model_dir: ./output/rec/test/
save_epoch_step: 3 save_epoch_step: 500
eval_batch_step: 2000 # evaluation is run every 5000 iterations after the 4000th iteration
train_batch_size_per_card: 256 eval_batch_step: 127
test_batch_size_per_card: 256 # if pretrained_model is saved in static mode, load_static_weights must set to True
image_shape: [3, 32, 100] load_static_weights: True
max_text_length: 25 cal_metric_during_train: True
character_type: en pretrained_model:
loss_type: ctc checkpoints: #output/rec/rec_crnn/best_accuracy
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir: save_inference_dir:
infer_img: use_visualdl: False
infer_img: doc/imgs_words/ch/word_1.jpg
# for data or label process
max_text_length: 80
character_dict_path: ppocr/utils/ppocr_keys_v1.txt
character_type: 'ch'
use_space_char: False
infer_mode: False
use_tps: False
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
Backbone: Optimizer:
function: ppocr.modeling.backbones.rec_mobilenet_v3,MobileNetV3 name: Adam
scale: 0.5 beta1: 0.9
model_name: large beta2: 0.999
learning_rate:
name: Cosine
lr: 0.001
warmup_epoch: 4
regularizer:
name: 'L2'
factor: 0.00001
Head: Architecture:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict type: rec
encoder_type: rnn algorithm: CRNN
SeqRNN: Transform:
Backbone:
name: MobileNetV3
scale: 0.5
model_name: small
small_stride: [ 1, 2, 2, 2 ]
Neck:
name: SequenceEncoder
encoder_type: fc
hidden_size: 96 hidden_size: 96
Head:
name: CTC
fc_decay: 0.00001
Loss: Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss name: CTCLoss
Optimizer: PostProcess:
function: ppocr.optimizer,AdamDecay name: CTCLabelDecode
base_lr: 0.001
beta1: 0.9 Metric:
beta2: 0.999 name: RecMetric
main_indicator: acc
TRAIN:
dataset:
name: SimpleDataSet
data_dir: /home/zhoujun20/rec
file_list:
- /home/zhoujun20/rec/real_data.txt # dataset1
ratio_list: [ 0.4,0.6 ]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecAug:
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader将按照此顺序返回list
loader:
batch_size: 256
shuffle: True
drop_last: True
num_workers: 6
EVAL:
dataset:
name: SimpleDataSet
data_dir: /home/zhoujun20/rec
file_list:
- /home/zhoujun20/rec/label_val_all.txt
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader将按照此顺序返回list
loader:
shuffle: False
drop_last: False
batch_size: 256
num_workers: 6

106
configs/rec/rec_mv3_none_bilstm_ctc_lmdb.yml
Unescape View File

@ -0,0 +1,106 @@
Global:
use_gpu: true
epoch_num: 500
log_smooth_window: 20
print_batch_step: 1
save_model_dir: ./output/rec/test/
save_epoch_step: 500
# evaluation is run every 5000 iterations after the 4000th iteration
eval_batch_step: 1016
# if pretrained_model is saved in static mode, load_static_weights must set to True
load_static_weights: True
cal_metric_during_train: True
pretrained_model:
checkpoints: #output/rec/rec_crnn/best_accuracy
save_inference_dir:
use_visualdl: True
infer_img: doc/imgs_words/ch/word_1.jpg
# for data or label process
max_text_length: 80
character_dict_path: /home/zhoujun20/rec/lmdb/dict.txt
character_type: 'ch'
use_space_char: True
infer_mode: False
use_tps: False
Optimizer:
name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
name: Cosine
lr: 0.0005
warmup_epoch: 1
regularizer:
name: 'L2'
factor: 0.00001
Architecture:
type: rec
algorithm: CRNN
Transform:
Backbone:
name: MobileNetV3
scale: 0.5
model_name: small
small_stride: [ 1, 2, 2, 2 ]
Neck:
name: SequenceEncoder
encoder_type: rnn
hidden_size: 48
Head:
name: CTC
fc_decay: 0.00001
Loss:
name: CTCLoss
PostProcess:
name: CTCLabelDecode
Metric:
name: RecMetric
main_indicator: acc
TRAIN:
dataset:
name: LMDBDateSet
file_list:
- /home/zhoujun20/rec/lmdb/train # dataset1
ratio_list: [ 0.4,0.6 ]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecAug:
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader将按照此顺序返回list
loader:
batch_size: 256
shuffle: True
drop_last: True
num_workers: 6
EVAL:
dataset:
name: LMDBDateSet
file_list:
- /home/zhoujun20/rec/lmdb/val
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader将按照此顺序返回list
loader:
shuffle: False
drop_last: False
batch_size: 256
num_workers: 6

41
configs/rec/rec_mv3_none_none_ctc.yml
Unescape View File

@ -1,41 +0,0 @@
Global:
algorithm: Rosetta
use_gpu: true
epoch_num: 72
log_smooth_window: 20
print_batch_step: 10
save_model_dir: output/rec_Rosetta
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 100]
max_text_length: 25
character_type: en
loss_type: ctc
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
Backbone:
function: ppocr.modeling.backbones.rec_mobilenet_v3,MobileNetV3
scale: 0.5
model_name: large
Head:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
encoder_type: reshape
Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999

54
configs/rec/rec_mv3_tps_bilstm_attn.yml
Unescape View File

@ -1,54 +0,0 @@
Global:
algorithm: RARE
use_gpu: true
epoch_num: 72
log_smooth_window: 20
print_batch_step: 10
save_model_dir: output/rec_RARE
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 100]
max_text_length: 25
character_type: en
loss_type: attention
tps: true
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
TPS:
function: ppocr.modeling.stns.tps,TPS
num_fiducial: 20
loc_lr: 0.1
model_name: small
Backbone:
function: ppocr.modeling.backbones.rec_mobilenet_v3,MobileNetV3
scale: 0.5
model_name: large
Head:
function: ppocr.modeling.heads.rec_attention_head,AttentionPredict
encoder_type: rnn
SeqRNN:
hidden_size: 96
Attention:
decoder_size: 96
word_vector_dim: 96
Loss:
function: ppocr.modeling.losses.rec_attention_loss,AttentionLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999

51
configs/rec/rec_mv3_tps_bilstm_ctc.yml
Unescape View File

@ -1,51 +0,0 @@
Global:
algorithm: STARNet
use_gpu: true
epoch_num: 72
log_smooth_window: 20
print_batch_step: 10
save_model_dir: output/rec_STARNet
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 100]
max_text_length: 25
character_type: en
loss_type: ctc
tps: true
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
TPS:
function: ppocr.modeling.stns.tps,TPS
num_fiducial: 20
loc_lr: 0.1
model_name: small
Backbone:
function: ppocr.modeling.backbones.rec_mobilenet_v3,MobileNetV3
scale: 0.5
model_name: large
Head:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
encoder_type: rnn
SeqRNN:
hidden_size: 96
Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999

127
configs/rec/rec_r34_vd_none_bilstm_ctc.yml
Unescape View File

@ -1,43 +1,106 @@
Global: Global:
algorithm: CRNN use_gpu: false
use_gpu: true epoch_num: 500
epoch_num: 72
log_smooth_window: 20 log_smooth_window: 20
print_batch_step: 10 print_batch_step: 10
save_model_dir: output/rec_CRNN save_model_dir: ./output/rec/test/
save_epoch_step: 3 save_epoch_step: 500
eval_batch_step: 2000 # evaluation is run every 5000 iterations after the 4000th iteration
train_batch_size_per_card: 256 eval_batch_step: 127
test_batch_size_per_card: 256 # if pretrained_model is saved in static mode, load_static_weights must set to True
image_shape: [3, 32, 100] load_static_weights: True
max_text_length: 25 cal_metric_during_train: True
character_type: en pretrained_model:
loss_type: ctc checkpoints: #output/rec/rec_crnn/best_accuracy
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir: save_inference_dir:
infer_img: use_visualdl: False
infer_img: doc/imgs_words/ch/word_1.jpg
# for data or label process
max_text_length: 80
character_dict_path: ppocr/utils/ppocr_keys_v1.txt
character_type: 'ch'
use_space_char: False
infer_mode: False
use_tps: False
Optimizer:
name: Adam
beta1: 0.9
beta2: 0.999
learning_rate:
name: Cosine
lr: 0.001
warmup_epoch: 4
regularizer:
name: 'L2'
factor: 0.00001
Architecture: Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel type: rec
algorithm: CRNN
Transform:
Backbone:
name: ResNet
layers: 200
Neck:
name: SequenceEncoder
encoder_type: fc
hidden_size: 96
Head:
name: CTC
fc_decay: 0.00001
Backbone: Loss:
function: ppocr.modeling.backbones.rec_resnet_vd,ResNet name: CTCLoss
layers: 34
Head: PostProcess:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict name: CTCLabelDecode
encoder_type: rnn
SeqRNN:
hidden_size: 256
Loss: Metric:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss name: RecMetric
main_indicator: acc
Optimizer: TRAIN:
function: ppocr.optimizer,AdamDecay dataset:
base_lr: 0.001 name: SimpleDataSet
beta1: 0.9 data_dir: /home/zhoujun20/rec
beta2: 0.999 file_list:
- /home/zhoujun20/rec/real_data.txt # dataset1
ratio_list: [ 0.4,0.6 ]
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecAug:
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader将按照此顺序返回list
loader:
batch_size: 256
shuffle: True
drop_last: True
num_workers: 6
EVAL:
dataset:
name: SimpleDataSet
data_dir: /home/zhoujun20/rec
file_list:
- /home/zhoujun20/rec/label_val_all.txt
transforms:
- DecodeImage: # load image
img_mode: BGR
channel_first: False
- CTCLabelEncode: # Class handling label
- RecResizeImg:
image_shape: [ 3,32,320 ]
- keepKeys:
keep_keys: [ 'image','label','length' ] # dataloader将按照此顺序返回list
loader:
shuffle: False
drop_last: False
batch_size: 256
num_workers: 6

40
configs/rec/rec_r34_vd_none_none_ctc.yml
Unescape View File

@ -1,40 +0,0 @@
Global:
algorithm: Rosetta
use_gpu: true
epoch_num: 72
log_smooth_window: 20
print_batch_step: 10
save_model_dir: output/rec_Rosetta
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 100]
max_text_length: 25
character_type: en
loss_type: ctc
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
Backbone:
function: ppocr.modeling.backbones.rec_resnet_vd,ResNet
layers: 34
Head:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
encoder_type: reshape
Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999

53
configs/rec/rec_r34_vd_tps_bilstm_attn.yml
Unescape View File

@ -1,53 +0,0 @@
Global:
algorithm: RARE
use_gpu: true
epoch_num: 72
log_smooth_window: 20
print_batch_step: 10
save_model_dir: output/rec_RARE
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 100]
max_text_length: 25
character_type: en
loss_type: attention
tps: true
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
TPS:
function: ppocr.modeling.stns.tps,TPS
num_fiducial: 20
loc_lr: 0.1
model_name: large
Backbone:
function: ppocr.modeling.backbones.rec_resnet_vd,ResNet
layers: 34
Head:
function: ppocr.modeling.heads.rec_attention_head,AttentionPredict
encoder_type: rnn
SeqRNN:
hidden_size: 256
Attention:
decoder_size: 128
word_vector_dim: 128
Loss:
function: ppocr.modeling.losses.rec_attention_loss,AttentionLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999

50
configs/rec/rec_r34_vd_tps_bilstm_ctc.yml
Unescape View File

@ -1,50 +0,0 @@
Global:
algorithm: STARNet
use_gpu: true
epoch_num: 72
log_smooth_window: 20
print_batch_step: 10
save_model_dir: output/rec_STARNet
save_epoch_step: 3
eval_batch_step: 2000
train_batch_size_per_card: 256
test_batch_size_per_card: 256
image_shape: [3, 32, 100]
max_text_length: 25
character_type: en
loss_type: ctc
tps: true
reader_yml: ./configs/rec/rec_benchmark_reader.yml
pretrain_weights:
checkpoints:
save_inference_dir:
infer_img:
Architecture:
function: ppocr.modeling.architectures.rec_model,RecModel
TPS:
function: ppocr.modeling.stns.tps,TPS
num_fiducial: 20
loc_lr: 0.1
model_name: large
Backbone:
function: ppocr.modeling.backbones.rec_resnet_vd,ResNet
layers: 34
Head:
function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
encoder_type: rnn
SeqRNN:
hidden_size: 256
Loss:
function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss
Optimizer:
function: ppocr.optimizer,AdamDecay
base_lr: 0.001
beta1: 0.9
beta2: 0.999

Some files were not shown because too many files have changed in this diff Show More

Write Preview
Loading…
Cancel
Save