robust_scanner_0805

lrr_ocr/lrr_SATRN/config/_base_/datasets/CASIA_HWDB_official_1x_data.py

@@ -0,0 +1,14 @@
+CASIA_HWDB_official_1x_data_textrecog_data_root = '/lirunrui/datasets/mmocr_CASIA_HWDB_official_1x_dataset'
+
+CASIA_HWDB_official_1x_data_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_official_1x_data_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+CASIA_HWDB_official_1x_data_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_official_1x_data_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_SATRN/config/satrn/satrn_shallow_5e_st_HWCR.py

@@ -1,6 +1,7 @@
 _base_ = [
     '../_base_/datasets/scut_hccdoc.py',
     '../_base_/datasets/CASIA_HWDB_chineseocr_data.py',
+    '../_base_/datasets/CASIA_HWDB_official_1x_data.py',
     '../_base_/datasets/CASIA_HWDB_official_2x_data.py',
     '../_base_/datasets/bnu_EnsExam_ppocrlabel.py',
     '../_base_/default_runtime.py',
@@ -14,12 +15,16 @@
     _base_.CASIA_HWDB_official_2x_data_textrecog_train,
     _base_.CASIA_HWDB_official_2x_data_textrecog_val,
     _base_.scut_hccdoc_textrecog_train,
+    # new add CASIA_HWDB_official_1x
+    _base_.CASIA_HWDB_official_1x_data_textrecog_train,
 ]
 
 test_list = [
     _base_.CASIA_HWDB_official_2x_data_textrecog_test,
     _base_.scut_hccdoc_textrecog_test,
     _base_.bnu_EnsExam_ppocrlabel_textrecog_test,
+    # new add CASIA_HWDB_official_1x
+    _base_.CASIA_HWDB_official_1x_data_textrecog_test,
 ]
 
 train_dataset = dict(
@@ -28,7 +33,7 @@
     type='ConcatDataset', datasets=test_list, pipeline=_base_.test_pipeline)
 
 default_hooks = dict(
-    logger=dict(type='LoggerHook', interval=1000), # 原本是100，但是batch_size设小的话那这个就输出的太多了。
+    logger=dict(type='LoggerHook', interval=2000), # 原本是100，但是batch_size设小的话那这个就输出的太多了。
     checkpoint=dict(type='CheckpointHook',
                     interval=1,
                     by_epoch=True, 
@@ -39,7 +44,7 @@
 )
 
 train_dataloader = dict(
-    batch_size=32,  # 原本是64，小一点。
+    batch_size=24,  # 原本是64，小一点。
     num_workers=32,
     persistent_workers=True,
     pin_memory=True,
@@ -67,7 +72,7 @@
         dict(type='CharMetric'),
         dict(type='CRandARMetric')
     ],
-    dataset_prefixes=['Casia_Hwdb_2x','SCUT_HccDoc','bnu_EnsExam_ppocrlabel'])   # 这个顺序要和 test_list 一致，要不然就乱了
+    dataset_prefixes=['Casia_Hwdb_2x','SCUT_HccDoc','bnu_EnsExam_ppocrlabel','Casia_Hwdb_1x'])   # 这个顺序要和 test_list 一致，要不然就乱了
 
 test_evaluator = val_evaluator
 

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/CASIA_HWDB_chineseocr_data.py

@@ -0,0 +1,14 @@
+CASIA_HWDB_chineseocr_data_textrecog_data_root = '/lirunrui/datasets/mmocr_CASIA_HWDB_chinese_ocr_dataset'
+
+CASIA_HWDB_chineseocr_data_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_chineseocr_data_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+CASIA_HWDB_chineseocr_data_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_chineseocr_data_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/CASIA_HWDB_official_1x_data.py

@@ -0,0 +1,14 @@
+CASIA_HWDB_official_1x_data_textrecog_data_root = '/lirunrui/datasets/mmocr_CASIA_HWDB_official_1x_dataset'
+
+CASIA_HWDB_official_1x_data_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_official_1x_data_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+CASIA_HWDB_official_1x_data_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_official_1x_data_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/CASIA_HWDB_official_2x_data.py

@@ -0,0 +1,21 @@
+CASIA_HWDB_official_2x_data_textrecog_data_root = '/lirunrui/datasets/mmocr_CASIA_HWDB_official_2x_dataset'
+
+CASIA_HWDB_official_2x_data_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_official_2x_data_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+CASIA_HWDB_official_2x_data_textrecog_val = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_official_2x_data_textrecog_data_root,
+    ann_file='textrecog_val.json',
+    # test_mode=True,   # 这玩意我也当做训练集扔进去
+    pipeline=None)
+
+CASIA_HWDB_official_2x_data_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=CASIA_HWDB_official_2x_data_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/bnu_EnsExam_ppocrlabel.py

@@ -0,0 +1,14 @@
+bnu_EnsExam_ppocrlabel_textrecog_data_root = '/lirunrui/datasets/mmocr_bnu_EnsExam_PPOCRLabel'
+
+bnu_EnsExam_ppocrlabel_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=bnu_EnsExam_ppocrlabel_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+bnu_EnsExam_ppocrlabel_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=bnu_EnsExam_ppocrlabel_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/coco_text_v1.py

@@ -0,0 +1,8 @@
+cocotextv1_textrecog_data_root = 'data/rec/coco_text_v1'
+
+cocotextv1_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=cocotextv1_textrecog_data_root,
+    ann_file='train_labels.json',
+    test_mode=False,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/cute80.py

@@ -0,0 +1,8 @@
+cute80_textrecog_data_root = 'data/cute80'
+
+cute80_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=cute80_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/icdar2011.py

@@ -0,0 +1,8 @@
+icdar2011_textrecog_data_root = 'data/rec/icdar_2011/'
+
+icdar2011_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=icdar2011_textrecog_data_root,
+    ann_file='train_labels.json',
+    test_mode=False,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/icdar2013.py

@@ -0,0 +1,21 @@
+icdar2013_textrecog_data_root = 'data/icdar2013'
+
+icdar2013_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=icdar2013_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+icdar2013_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=icdar2013_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)
+
+icdar2013_857_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=icdar2013_textrecog_data_root,
+    ann_file='textrecog_test_857.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/icdar2015.py

@@ -0,0 +1,21 @@
+icdar2015_textrecog_data_root = 'data/icdar2015'
+
+icdar2015_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=icdar2015_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+icdar2015_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=icdar2015_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)
+
+icdar2015_1811_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=icdar2015_textrecog_data_root,
+    ann_file='textrecog_test_1811.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/iiit5k.py

@@ -0,0 +1,14 @@
+iiit5k_textrecog_data_root = 'data/iiit5k'
+
+iiit5k_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=iiit5k_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+iiit5k_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=iiit5k_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/mjsynth.py

@@ -0,0 +1,13 @@
+mjsynth_textrecog_data_root = 'data/mjsynth'
+
+mjsynth_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=mjsynth_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+mjsynth_sub_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=mjsynth_textrecog_data_root,
+    ann_file='subset_textrecog_train.json',
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/scut_hccdoc.py

@@ -0,0 +1,24 @@
+scut_hccdoc_textrecog_data_root = '/lirunrui/datasets/mmocr_SCUT_HCCDoc_Dataset'
+default_hooks = dict(
+
+    visualization=dict(
+        type='VisualizationHook',
+        interval=1,
+        enable=False,
+        show=False,
+        draw_gt=False,
+        draw_pred=False,
+        font_properties='/usr/share/fonts/fonts_library/simsun.ttc'),
+)
+scut_hccdoc_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=scut_hccdoc_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+scut_hccdoc_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=scut_hccdoc_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/svt.py

@@ -0,0 +1,14 @@
+svt_textrecog_data_root = 'data/svt'
+
+svt_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=svt_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+svt_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=svt_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/svtp.py

@@ -0,0 +1,14 @@
+svtp_textrecog_data_root = 'data/svtp'
+
+svtp_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=svtp_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+svtp_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=svtp_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/synthtext.py

@@ -0,0 +1,19 @@
+synthtext_textrecog_data_root = '/lirunrui/datasets/synthtext'
+
+synthtext_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=synthtext_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    pipeline=None)
+
+synthtext_sub_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=synthtext_textrecog_data_root,
+    ann_file='subset_textrecog_train.json',
+    pipeline=None)
+
+synthtext_an_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=synthtext_textrecog_data_root,
+    ann_file='alphanumeric_textrecog_train.json',
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/synthtext_add.py

@@ -0,0 +1,8 @@
+synthtext_add_textrecog_data_root = 'data/rec/synthtext_add/'
+
+synthtext_add_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=synthtext_add_textrecog_data_root,
+    ann_file='train_labels.json',
+    test_mode=False,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/totaltext.py

@@ -0,0 +1,15 @@
+totaltext_textrecog_data_root = 'data/totaltext/'
+
+totaltext_textrecog_train = dict(
+    type='OCRDataset',
+    data_root=totaltext_textrecog_data_root,
+    ann_file='textrecog_train.json',
+    test_mode=False,
+    pipeline=None)
+
+totaltext_textrecog_test = dict(
+    type='OCRDataset',
+    data_root=totaltext_textrecog_data_root,
+    ann_file='textrecog_test.json',
+    test_mode=True,
+    pipeline=None)

lrr_ocr/lrr_robust_scanner/configs/_base_/datasets/toy_data.py

@@ -0,0 +1,17 @@
+toy_data_root = 'tests/data/rec_toy_dataset/'
+
+toy_rec_train = dict(
+    type='OCRDataset',
+    data_root=toy_data_root,
+    data_prefix=dict(img_path='imgs/'),
+    ann_file='labels.json',
+    pipeline=None,
+    test_mode=False)
+
+toy_rec_test = dict(
+    type='OCRDataset',
+    data_root=toy_data_root,
+    data_prefix=dict(img_path='imgs/'),
+    ann_file='labels.json',
+    pipeline=None,
+    test_mode=True)