mirrors
/
mmocr
mirror of https://github.com/open-mmlab/mmocr.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

[Enhancement] Deprecate file_client_args and use backend_args instead (#1765) 

* tmp commit

* remove
pull/1786/head
Tong Gao 2023-03-20 10:33:20 +08:00 committed by GitHub
parent 63a6ed4e6c
commit 9caacc76ee
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
39 changed files with 144 additions and 300 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
configs/textdet/dbnet/_base_dbnet_resnet18_fpnc.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
model = dict(
type='DBNet',
backbone=dict(
@ -27,10 +25,7 @@ model = dict(
pad_size_divisor=32))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -55,10 +50,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1333, 736), keep_ratio=True),
dict(
type='LoadOCRAnnotations',

12
configs/textdet/dbnet/_base_dbnet_resnet50-dcnv2_fpnc.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
model = dict(
type='DBNet',
backbone=dict(
@ -29,10 +27,7 @@ model = dict(
pad_size_divisor=32))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_bbox=True,
@ -57,10 +52,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(4068, 1024), keep_ratio=True),
dict(
type='LoadOCRAnnotations',

7
configs/textdet/dbnet/dbnet_resnet18_fpnc_100k_synthtext.py
View File

@ -5,13 +5,8 @@ _base_ = [
'../_base_/schedules/schedule_sgd_100k.py',
]
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,

12
configs/textdet/dbnet/dbnet_resnet18_fpnc_1200e_totaltext.py
View File

@ -5,13 +5,8 @@ _base_ = [
'../_base_/schedules/schedule_sgd_1200e.py',
]
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -37,10 +32,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1333, 736), keep_ratio=True),
dict(
type='LoadOCRAnnotations',

12
configs/textdet/drrg/_base_drrg_resnet50_fpn-unet.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
model = dict(
type='DRRG',
backbone=dict(
@ -29,10 +27,7 @@ model = dict(
pad_size_divisor=32))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_bbox=True,
@ -82,10 +77,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1024, 640), keep_ratio=True),
# add loading annotation after ``Resize`` because ground truth
# does not need to do resize data transform

12
configs/textdet/fcenet/_base_fcenet_resnet50_fpn.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
model = dict(
type='FCENet',
backbone=dict(
@ -41,10 +39,7 @@ model = dict(
pad_size_divisor=32))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -96,10 +91,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(2260, 2260), keep_ratio=True),
# add loading annotation after ``Resize`` because ground truth
# does not need to do resize data transform

6
configs/textdet/fcenet/fcenet_resnet50-dcnv2_fpn_1500e_ctw1500.py
View File

@ -12,17 +12,13 @@ param_scheduler = [
dict(type='PolyLR', power=0.9, eta_min=1e-7, end=1500),
]
file_client_args = dict(backend='disk')
# dataset settings
ctw1500_textdet_train = _base_.ctw1500_textdet_train
ctw1500_textdet_test = _base_.ctw1500_textdet_test
# test pipeline for CTW1500
ctw_test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1080, 736), keep_ratio=True),
# add loading annotation after ``Resize`` because ground truth
# does not need to do resize data transform

12
configs/textdet/maskrcnn/_base_mask-rcnn_resnet50_fpn.py
View File

@ -1,7 +1,5 @@
_base_ = ['mmdet::_base_/models/mask-rcnn_r50_fpn.py']
file_client_args = dict(backend='disk')
mask_rcnn = _base_.pop('model')
# Adapt Mask R-CNN model to OCR task
mask_rcnn.update(
@ -18,10 +16,7 @@ mask_rcnn.update(
model = dict(type='MMDetWrapper', text_repr_type='poly', cfg=mask_rcnn)
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -49,10 +44,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1920, 1920), keep_ratio=True),
dict(
type='LoadOCRAnnotations',

5
configs/textdet/maskrcnn/mask-rcnn_resnet50_fpn_160e_ctw1500.py
View File

@ -20,10 +20,7 @@ ctw1500_textdet_test = _base_.ctw1500_textdet_test
# test pipeline for CTW1500
ctw_test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk'),
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1600, 1600), keep_ratio=True),
# add loading annotation after ``Resize`` because ground truth
# does not need to do resize data transform

11
configs/textdet/panet/_base_panet_resnet18_fpem-ffm.py
View File

@ -32,12 +32,8 @@ model = dict(
),
postprocessor=dict(type='PANPostprocessor', text_repr_type='quad')))
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -60,10 +56,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
# TODO Replace with mmcv.RescaleToShort when it's ready
dict(
type='ShortScaleAspectJitter',

11
configs/textdet/panet/panet_resnet18_fpem-ffm_600e_ctw1500.py
View File

@ -9,12 +9,8 @@ model = dict(det_head=dict(module_loss=dict(shrink_ratio=(1, 0.7))))
default_hooks = dict(checkpoint=dict(type='CheckpointHook', interval=20), )
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -37,10 +33,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
# TODO Replace with mmcv.RescaleToShort when it's ready
dict(
type='ShortScaleAspectJitter',

11
configs/textdet/panet/panet_resnet50_fpem-ffm_600e_icdar2017.py
View File

@ -7,12 +7,8 @@ _base_ = [
default_hooks = dict(checkpoint=dict(type='CheckpointHook', interval=20), )
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -35,10 +31,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
# TODO Replace with mmcv.RescaleToShort when it's ready
dict(
type='ShortScaleAspectJitter',

12
configs/textdet/psenet/_base_psenet_resnet50_fpnf.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
model = dict(
type='PSENet',
backbone=dict(
@ -32,10 +30,7 @@ model = dict(
pad_size_divisor=32))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,
@ -58,10 +53,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(2240, 2240), keep_ratio=True),
dict(
type='LoadOCRAnnotations',

5
configs/textdet/psenet/psenet_resnet50_fpnf_600e_ctw1500.py
View File

@ -17,10 +17,7 @@ ctw1500_textdet_train = _base_.ctw1500_textdet_train
ctw1500_textdet_test = _base_.ctw1500_textdet_test
test_pipeline_ctw = [
dict(
type='LoadImageFromFile',
file_client_args=_base_.file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1280, 1280), keep_ratio=True),
dict(
type='LoadOCRAnnotations',

12
configs/textdet/textsnake/_base_textsnake_resnet50_fpn-unet.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
model = dict(
type='TextSnake',
backbone=dict(
@ -28,10 +26,7 @@ model = dict(
pad_size_divisor=32))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_bbox=True,
@ -72,10 +67,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(1333, 736), keep_ratio=True),
# add loading annotation after ``Resize`` because ground truth
# does not need to do resize data transform

12
configs/textrecog/abinet/_base_abinet-vision.py
View File

@ -39,14 +39,8 @@ model = dict(
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375]))
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(type='Resize', scale=(128, 32)),
dict(
@ -107,7 +101,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(type='Resize', scale=(128, 32)),
# add loading annotation after ``Resize`` because ground truth
# does not need to do resize data transform
@ -118,7 +112,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[

12
configs/textrecog/aster/_base_aster.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
dictionary = dict(
type='Dictionary',
dict_file='{{ fileDirname }}/../../../dicts/english_digits_symbols.txt',
@ -48,11 +46,7 @@ model = dict(
std=[127.5, 127.5, 127.5]))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=5),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=5),
dict(type='LoadOCRAnnotations', with_text=True),
dict(type='Resize', scale=(256, 64)),
dict(
@ -61,7 +55,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(type='Resize', scale=(256, 64)),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
@ -71,7 +65,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[[

12
configs/textrecog/crnn/_base_crnn_mini-vgg.py
View File

@ -18,12 +18,10 @@ model = dict(
data_preprocessor=dict(
type='TextRecogDataPreprocessor', mean=[127], std=[127]))
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
color_type='grayscale',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
@ -34,10 +32,7 @@ train_pipeline = [
]
test_pipeline = [
dict(
type='LoadImageFromFile',
color_type='grayscale',
file_client_args=file_client_args),
dict(type='LoadImageFromFile', color_type='grayscale'),
dict(
type='RescaleToHeight',
height=32,
@ -53,10 +48,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(
type='LoadImageFromFile',
color_type='grayscale',
file_client_args=file_client_args),
dict(type='LoadImageFromFile', color_type='grayscale'),
dict(
type='TestTimeAug',
transforms=[

12
configs/textrecog/master/_base_master_resnet31.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
dictionary = dict(
type='Dictionary',
dict_file='{{ fileDirname }}/../../../dicts/english_digits_symbols.txt',
@ -75,11 +73,7 @@ model = dict(
std=[127.5, 127.5, 127.5]))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='RescaleToHeight',
@ -94,7 +88,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='RescaleToHeight',
height=48,
@ -111,7 +105,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[

12
configs/textrecog/nrtr/_base_nrtr_modality-transform.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
dictionary = dict(
type='Dictionary',
dict_file='{{ fileDirname }}/../../../dicts/english_digits_symbols.txt',
@ -26,11 +24,7 @@ model = dict(
std=[58.395, 57.12, 57.375]))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='RescaleToHeight',
@ -45,7 +39,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='RescaleToHeight',
height=32,
@ -62,7 +56,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[

12
configs/textrecog/nrtr/_base_nrtr_resnet31.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
dictionary = dict(
type='Dictionary',
dict_file='{{ fileDirname }}/../../../dicts/english_digits_symbols.txt',
@ -32,11 +30,7 @@ model = dict(
std=[58.395, 57.12, 57.375]))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='RescaleToHeight',
@ -51,7 +45,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='RescaleToHeight',
height=32,
@ -68,7 +62,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[

12
configs/textrecog/robust_scanner/_base_robustscanner_resnet31.py
View File

@ -29,14 +29,8 @@ model = dict(
dictionary=dictionary,
max_seq_len=30))
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='RescaleToHeight',
@ -51,7 +45,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='RescaleToHeight',
height=48,
@ -68,7 +62,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[

11
configs/textrecog/sar/_base_sar_resnet31_parallel-decoder.py
View File

@ -35,13 +35,8 @@ model = dict(
dictionary=dictionary,
max_seq_len=30))
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='RescaleToHeight',
@ -56,7 +51,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='RescaleToHeight',
height=48,
@ -73,7 +68,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[

12
configs/textrecog/satrn/_base_satrn_shallow.py
View File

@ -1,5 +1,3 @@
file_client_args = dict(backend='disk')
dictionary = dict(
type='Dictionary',
dict_file='{{ fileDirname }}/../../../dicts/english_digits_symbols.txt',
@ -42,11 +40,7 @@ model = dict(
std=[58.395, 57.12, 57.375]))
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(type='Resize', scale=(100, 32), keep_ratio=False),
dict(
@ -55,7 +49,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(type='Resize', scale=(100, 32), keep_ratio=False),
# add loading annotation after ``Resize`` because ground truth
# does not need to do resize data transform
@ -66,7 +60,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[

12
configs/textrecog/svtr/_base_svtr-tiny.py
View File

@ -37,14 +37,8 @@ model = dict(
data_preprocessor=dict(
type='TextRecogDataPreprocessor', mean=[127.5], std=[127.5]))
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
ignore_empty=True,
min_size=5),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=5),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='RandomApply',
@ -117,7 +111,7 @@ train_pipeline = [
]
test_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(type='Resize', scale=(256, 64)),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
@ -126,7 +120,7 @@ test_pipeline = [
]
tta_pipeline = [
dict(type='LoadImageFromFile', file_client_args=file_client_args),
dict(type='LoadImageFromFile'),
dict(
type='TestTimeAug',
transforms=[[

1
docs/en/basic_concepts/transforms.md
View File

@ -24,7 +24,6 @@ train_pipeline_r18 = [
# Loading images
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
# Loading annotations
dict(

5
docs/en/user_guides/config.md
View File

@ -332,12 +332,10 @@ Due to the specificity of the OCR task, different models have different data aug
```Python
# Data Augmentation
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
color_type='grayscale',
file_client_args=dict(backend='disk'),
ignore_empty=True,
min_size=5),
dict(type='LoadOCRAnnotations', with_text=True),
@ -349,8 +347,7 @@ train_pipeline = [
test_pipeline = [
dict(
type='LoadImageFromFile',
color_type='grayscale',
file_client_args=dict(backend='disk')),
color_type='grayscale'),
dict(
type='RescaleToHeight',
height=32,

1
docs/zh_cn/basic_concepts/transforms.md
View File

@ -24,7 +24,6 @@ train_pipeline_r18 = [
# 数据读取（图像）
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
# 数据读取（标注）
dict(

5
docs/zh_cn/user_guides/config.md
View File

@ -329,12 +329,10 @@ MMOCR 中，数据集的构建与数据准备是相互解耦的。也就是说
```Python
# 数据增强
file_client_args = dict(backend='disk')
train_pipeline = [
dict(
type='LoadImageFromFile',
color_type='grayscale',
file_client_args=dict(backend='disk'),
ignore_empty=True,
min_size=5),
dict(type='LoadOCRAnnotations', with_text=True),
@ -346,8 +344,7 @@ train_pipeline = [
test_pipeline = [
dict(
type='LoadImageFromFile',
color_type='grayscale',
file_client_args=dict(backend='disk')),
color_type='grayscale'),
dict(
type='RescaleToHeight',
height=32,

13
mmocr/datasets/recog_text_dataset.py
View File

@ -7,8 +7,6 @@ from mmengine.fileio import list_from_file
from mmocr.registry import DATASETS, TASK_UTILS
# TODO: replace all list_from_file from mmengine
@DATASETS.register_module()
class RecogTextDataset(BaseDataset):
@ -33,9 +31,8 @@ class RecogTextDataset(BaseDataset):
Args:
ann_file (str): Annotation file path. Defaults to ''.
file_client_args (dict, optional): Arguments to instantiate a
FileClient. See :class:`mmengine.fileio.FileClient` for details.
Default: None.
backend_args (dict, optional): Arguments to instantiate the
prefix of uri corresponding backend. Defaults to None.
parse_cfg (dict, optional): Config of parser for parsing annotations.
Use ``LineJsonParser`` when the annotation file is in jsonl format
with keys of ``filename`` and ``text``. The keys in parse_cfg
@ -75,7 +72,7 @@ class RecogTextDataset(BaseDataset):
def __init__(self,
ann_file: str = '',
file_client_args=None,
backend_args=None,
parser_cfg: Optional[dict] = dict(
type='LineJsonParser', keys=['filename', 'text']),
metainfo: Optional[dict] = None,
@ -90,7 +87,7 @@ class RecogTextDataset(BaseDataset):
max_refetch: int = 1000) -> None:
self.parser = TASK_UTILS.build(parser_cfg)
self.file_client_args = file_client_args
self.backend_args = backend_args
super().__init__(
ann_file=ann_file,
metainfo=metainfo,
@ -112,7 +109,7 @@ class RecogTextDataset(BaseDataset):
"""
data_list = []
raw_anno_infos = list_from_file(
self.ann_file, file_client_args=self.file_client_args)
self.ann_file, backend_args=self.backend_args)
for raw_anno_info in raw_anno_infos:
data_list.append(self.parse_data_info(raw_anno_info))
return data_list

69
mmocr/datasets/transforms/loading.py
View File

@ -4,7 +4,7 @@ import warnings
from typing import Optional, Union
import mmcv
import mmengine
import mmengine.fileio as fileio
import numpy as np
from mmcv.transforms import BaseTransform
from mmcv.transforms import LoadAnnotations as MMCV_LoadAnnotations
@ -39,7 +39,16 @@ class LoadImageFromFile(MMCV_LoadImageFromFile):
Defaults to 'cv2'.
file_client_args (dict): Arguments to instantiate a FileClient.
See :class:`mmengine.fileio.FileClient` for details.
Defaults to ``dict(backend='disk')``.
Defaults to None. It will be deprecated in future. Please use
``backend_args`` instead.
Deprecated in version 1.0.0rc6.
backend_args (dict, optional): Instantiates the corresponding file
backend. It may contain `backend` key to specify the file
backend. If it contains, the file backend corresponding to this
value will be used and initialized with the remaining values,
otherwise the corresponding file backend will be selected
based on the prefix of the file path. Defaults to None.
New in version 1.0.0rc6.
ignore_empty (bool): Whether to allow loading empty image or file path
not existent. Defaults to False.
min_size (int): The minimum size of the image to be loaded. If the
@ -47,20 +56,36 @@ class LoadImageFromFile(MMCV_LoadImageFromFile):
broken image. Defaults to 0.
"""
def __init__(self,
to_float32: bool = False,
color_type: str = 'color',
imdecode_backend: str = 'cv2',
file_client_args: dict = dict(backend='disk'),
min_size: int = 0,
ignore_empty: bool = False) -> None:
def __init__(
self,
to_float32: bool = False,
color_type: str = 'color',
imdecode_backend: str = 'cv2',
file_client_args: Optional[dict] = None,
min_size: int = 0,
ignore_empty: bool = False,
*,
backend_args: Optional[dict] = None,
) -> None:
self.ignore_empty = ignore_empty
self.to_float32 = to_float32
self.color_type = color_type
self.imdecode_backend = imdecode_backend
self.file_client_args = file_client_args.copy()
self.file_client = mmengine.FileClient(**self.file_client_args)
self.min_size = min_size
self.file_client_args = file_client_args
self.backend_args = backend_args
if file_client_args is not None:
warnings.warn(
'"file_client_args" will be deprecated in future. '
'Please use "backend_args" instead', DeprecationWarning)
if backend_args is not None:
raise ValueError(
'"file_client_args" and "backend_args" cannot be set '
'at the same time.')
self.file_client_args = file_client_args.copy()
if backend_args is not None:
self.backend_args = backend_args.copy()
def transform(self, results: dict) -> Optional[dict]:
"""Functions to load image.
@ -74,7 +99,13 @@ class LoadImageFromFile(MMCV_LoadImageFromFile):
filename = results['img_path']
try:
img_bytes = self.file_client.get(filename)
if getattr(self, 'file_client_args', None) is not None:
file_client = fileio.FileClient.infer_client(
self.file_client_args, filename)
img_bytes = file_client.get(filename)
else:
img_bytes = fileio.get(
filename, backend_args=self.backend_args)
img = mmcv.imfrombytes(
img_bytes, flag=self.color_type, backend=self.imdecode_backend)
except Exception as e:
@ -103,8 +134,12 @@ class LoadImageFromFile(MMCV_LoadImageFromFile):
f'min_size={self.min_size}, '
f'to_float32={self.to_float32}, '
f"color_type='{self.color_type}', "
f"imdecode_backend='{self.imdecode_backend}', "
f'file_client_args={self.file_client_args})')
f"imdecode_backend='{self.imdecode_backend}', ")
if self.file_client_args is not None:
repr_str += f'file_client_args={self.file_client_args})'
else:
repr_str += f'backend_args={self.backend_args})'
return repr_str
@ -371,7 +406,11 @@ class LoadOCRAnnotations(MMCV_LoadAnnotations):
repr_str += f'with_polygon={self.with_polygon}, '
repr_str += f'with_text={self.with_text}, '
repr_str += f"imdecode_backend='{self.imdecode_backend}', "
repr_str += f'file_client_args={self.file_client_args})'
if self.file_client_args is not None:
repr_str += f'file_client_args={self.file_client_args})'
else:
repr_str += f'backend_args={self.backend_args})'
return repr_str

29
mmocr/engine/hooks/visualization_hook.py
View File

@ -1,9 +1,9 @@
# Copyright (c) OpenMMLab. All rights reserved.
import os.path as osp
from typing import Sequence, Union
from typing import Optional, Sequence, Union
import mmcv
import mmengine
import mmengine.fileio as fileio
from mmengine.hooks import Hook
from mmengine.runner import Runner
from mmengine.visualization import Visualizer
@ -26,9 +26,12 @@ class VisualizationHook(Hook):
show (bool): Whether to display the drawn image. Defaults to False.
wait_time (float): The interval of show in seconds. Defaults
to 0.
file_client_args (dict): Arguments to instantiate a FileClient.
See :class:`mmengine.fileio.FileClient` for details.
Defaults to ``dict(backend='disk')``.
backend_args (dict, optional): Instantiates the corresponding file
backend. It may contain `backend` key to specify the file
backend. If it contains, the file backend corresponding to this
value will be used and initialized with the remaining values,
otherwise the corresponding file backend will be selected
based on the prefix of the file path. Defaults to None.
"""
def __init__(
@ -40,7 +43,7 @@ class VisualizationHook(Hook):
draw_pred: bool = False,
draw_gt: bool = False,
wait_time: float = 0.,
file_client_args: dict = dict(backend='disk')
backend_args: Optional[dict] = None,
) -> None:
self._visualizer: Visualizer = Visualizer.get_current_instance()
self.interval = interval
@ -49,8 +52,7 @@ class VisualizationHook(Hook):
self.draw_pred = draw_pred
self.draw_gt = draw_gt
self.wait_time = wait_time
self.file_client_args = file_client_args.copy()
self.file_client = None
self.backend_args = backend_args
self.enable = enable
# TODO after MultiDatasetWrapper, rewrites this function and try to merge
@ -72,9 +74,6 @@ class VisualizationHook(Hook):
if self.enable is False:
return
if self.file_client is None:
self.file_client = mmengine.FileClient(**self.file_client_args)
# There is no guarantee that the same batch of images
# is visualized for each evaluation.
total_curr_iter = runner.iter + batch_idx
@ -83,7 +82,8 @@ class VisualizationHook(Hook):
if total_curr_iter % self.interval == 0:
for output in outputs:
img_path = output.img_path
img_bytes = self.file_client.get(img_path)
img_bytes = fileio.get(
img_path, backend_args=self.backend_args)
img = mmcv.imfrombytes(img_bytes, channel_order='rgb')
self._visualizer.add_datasample(
osp.splitext(osp.basename(img_path))[0],
@ -113,12 +113,9 @@ class VisualizationHook(Hook):
if self.enable is False:
return
if self.file_client is None:
self.file_client = mmengine.FileClient(**self.file_client_args)
for output in outputs:
img_path = output.img_path
img_bytes = self.file_client.get(img_path)
img_bytes = fileio.get(img_path, backend_args=self.backend_args)
img = mmcv.imfrombytes(img_bytes, channel_order='rgb')
self._visualizer.add_datasample(

3
mmocr/models/textrecog/recognizers/encoder_decoder_recognizer_tta.py
View File

@ -20,8 +20,7 @@ class EncoderDecoderRecognizerTTAModel(BaseTTAModel):
>>> tta_pipeline = [
>>> dict(
>>> type='LoadImageFromFile',
>>> color_type='grayscale',
>>> file_client_args=file_client_args),
>>> color_type='grayscale'),
>>> dict(
>>> type='TestTimeAug',
>>> transforms=[

11
projects/ABCNet/config/abcnet/_base_abcnet_resnet50_fpn.py
View File

@ -1,4 +1,3 @@
file_client_args = dict(backend='disk')
num_classes = 1
strides = [8, 16, 32, 64, 128]
bbox_coder = dict(type='mmdet.DistancePointBBoxCoder')
@ -126,10 +125,7 @@ model = dict(
))
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(2000, 4000), keep_ratio=True, backend='pillow'),
dict(
type='LoadOCRAnnotations',
@ -143,10 +139,7 @@ test_pipeline = [
]
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
with_polygon=True,

6
projects/ABCNet/config/abcnet_v2/_base_abcnet-v2_resnet50_bifpn.py
View File

@ -1,4 +1,3 @@
file_client_args = dict(backend='disk')
num_classes = 1
strides = [8, 16, 32, 64, 128]
bbox_coder = dict(type='mmdet.DistancePointBBoxCoder')
@ -105,10 +104,7 @@ model = dict(
))
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(type='Resize', scale=(2000, 4000), keep_ratio=True, backend='pillow'),
dict(
type='LoadOCRAnnotations',

12
projects/SPTS/config/spts/_base_spts_resnet50.py
View File

@ -1,8 +1,6 @@
custom_imports = dict(
imports=['projects.SPTS.spts'], allow_failed_imports=False)
file_client_args = dict(backend='disk')
dictionary = dict(
type='SPTSDictionary',
dict_file='{{ fileDirname }}/../../dicts/spts.txt',
@ -53,10 +51,7 @@ model = dict(
postprocessor=dict(type='SPTSPostprocessor', num_bins=num_bins)))
test_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
# dict(type='Resize', scale=(1000, 1824), keep_ratio=True),
dict(
type='RescaleToShortSide',
@ -73,10 +68,7 @@ test_pipeline = [
]
train_pipeline = [
dict(
type='LoadImageFromFile',
file_client_args=file_client_args,
color_type='color_ignore_orientation'),
dict(type='LoadImageFromFile', color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotationsWithBezier',
with_bbox=True,

7
projects/SPTS/spts/datasets/transforms/spts_transforms.py
View File

@ -137,7 +137,12 @@ class LoadOCRAnnotationsWithBezier(LoadOCRAnnotations):
repr_str += f'with_bezier={self.with_bezier}, '
repr_str += f'with_text={self.with_text}, '
repr_str += f"imdecode_backend='{self.imdecode_backend}', "
repr_str += f'file_client_args={self.file_client_args})'
if self.file_client_args is not None:
repr_str += f'file_client_args={self.file_client_args})'
else:
repr_str += f'backend_args={self.backend_args})'
return repr_str

4
tests/test_datasets/test_transforms/test_loading.py
View File

@ -29,7 +29,7 @@ class TestLoadImageFromFile(TestCase):
repr(transform),
('LoadImageFromFile(ignore_empty=False, min_size=0, '
"to_float32=False, color_type='color', imdecode_backend='cv2', "
"file_client_args={'backend': 'disk'})"))
'backend_args=None)'))
# to_float32
transform = LoadImageFromFile(to_float32=True)
@ -130,7 +130,7 @@ class TestLoadOCRAnnotations(TestCase):
repr(transform),
('LoadOCRAnnotations(with_bbox=True, with_label=True, '
'with_polygon=True, with_text=True, '
"imdecode_backend='cv2', file_client_args=None)"))
"imdecode_backend='cv2', backend_args=None)"))
class TestLoadKIEAnnotations(TestCase):

7
tools/analysis_tools/browse_dataset.py
View File

@ -259,7 +259,6 @@ def obtain_dataset_cfg(cfg: Config, phase: str, mode: str, task: str) -> Tuple:
pipeline=[
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk'),
color_type='color_ignore_orientation'),
dict(
type='LoadOCRAnnotations',
@ -277,11 +276,7 @@ def obtain_dataset_cfg(cfg: Config, phase: str, mode: str, task: str) -> Tuple:
name='visualizer',
vis_backends=[dict(type='LocalVisBackend')]),
pipeline=[
dict(
type='LoadImageFromFile',
file_client_args=dict(backend='disk'),
ignore_empty=True,
min_size=2),
dict(type='LoadImageFromFile', ignore_empty=True, min_size=2),
dict(type='LoadOCRAnnotations', with_text=True),
dict(
type='PackTextRecogInputs',