Update weights - 40e

model/text-det/psenet.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d42af4c913d9af97e9beeed47185fac769bcf65b7a653195e22340748d9eb335
-size 352719845
+oid sha256:ae9fd081c8004a7a8a6f3d1bab370637a819b8dcefe0d9c23c54c2d2502339aa
+size 353251813

model/text-det/psenet.py

@@ -29,114 +29,6 @@ model = dict(
         std=[58.395, 57.12, 57.375],
         bgr_to_rgb=True,
         pad_size_divisor=32))
-train_pipeline = [
-    dict(
-        type='LoadImageFromFile',
-        file_client_args=dict(backend='disk'),
-        color_type='color_ignore_orientation'),
-    dict(
-        type='LoadOCRAnnotations',
-        with_polygon=True,
-        with_bbox=True,
-        with_label=True),
-    dict(
-        type='TorchVisionWrapper',
-        op='ColorJitter',
-        brightness=0.12549019607843137,
-        saturation=0.5),
-    dict(type='FixInvalidPolygon'),
-    dict(type='ShortScaleAspectJitter', short_size=736, scale_divisor=32),
-    dict(type='RandomFlip', prob=0.5, direction='horizontal'),
-    dict(type='RandomRotate', max_angle=10),
-    dict(type='TextDetRandomCrop', target_size=(736, 736)),
-    dict(type='Pad', size=(736, 736)),
-    dict(
-        type='PackTextDetInputs',
-        meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
-]
-test_pipeline = [
-    dict(
-        type='LoadImageFromFile',
-        file_client_args=dict(backend='disk'),
-        color_type='color_ignore_orientation'),
-    dict(type='Resize', scale=(2240, 2240), keep_ratio=True),
-    dict(
-        type='LoadOCRAnnotations',
-        with_polygon=True,
-        with_bbox=True,
-        with_label=True),
-    dict(
-        type='PackTextDetInputs',
-        meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
-]
-thvc_textdet_data_root = 'data/det/vl+vc-textdet'
-thvc_textdet_train = dict(
-    type='OCRDataset',
-    data_root='data/det/vl+vc-textdet',
-    ann_file='textdet_train.json',
-    data_prefix=dict(img_path='imgs/'),
-    filter_cfg=dict(filter_empty_gt=True, min_size=32),
-    pipeline=[
-        dict(
-            type='LoadImageFromFile',
-            file_client_args=dict(backend='disk'),
-            color_type='color_ignore_orientation'),
-        dict(
-            type='LoadOCRAnnotations',
-            with_polygon=True,
-            with_bbox=True,
-            with_label=True),
-        dict(
-            type='TorchVisionWrapper',
-            op='ColorJitter',
-            brightness=0.12549019607843137,
-            saturation=0.5),
-        dict(type='FixInvalidPolygon'),
-        dict(type='ShortScaleAspectJitter', short_size=736, scale_divisor=32),
-        dict(type='RandomFlip', prob=0.5, direction='horizontal'),
-        dict(type='RandomRotate', max_angle=10),
-        dict(type='TextDetRandomCrop', target_size=(736, 736)),
-        dict(type='Pad', size=(736, 736)),
-        dict(
-            type='PackTextDetInputs',
-            meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
-    ])
-thvc_textdet_test = dict(
-    type='OCRDataset',
-    data_root='data/det/vl+vc-textdet',
-    ann_file='textdet_test.json',
-    data_prefix=dict(img_path='imgs/'),
-    test_mode=True,
-    pipeline=None)
-thvote_textdet_data_root = 'data/det/textdet-thvote'
-thvote_textdet_train = dict(
-    type='OCRDataset',
-    data_root='data/det/textdet-thvote',
-    ann_file='textdet_train.json',
-    data_prefix=dict(img_path='imgs/'),
-    filter_cfg=dict(filter_empty_gt=True, min_size=32),
-    pipeline=None)
-thvote_textdet_test = dict(
-    type='OCRDataset',
-    data_root='data/det/textdet-thvote',
-    ann_file='textdet_test.json',
-    data_prefix=dict(img_path='imgs/'),
-    test_mode=True,
-    pipeline=[
-        dict(
-            type='LoadImageFromFile',
-            file_client_args=dict(backend='disk'),
-            color_type='color_ignore_orientation'),
-        dict(type='Resize', scale=(2240, 2240), keep_ratio=True),
-        dict(
-            type='LoadOCRAnnotations',
-            with_polygon=True,
-            with_bbox=True,
-            with_label=True),
-        dict(
-            type='PackTextDetInputs',
-            meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
-    ])
 default_scope = 'mmocr'
 env_cfg = dict(
     cudnn_benchmark=True,
@@ -168,65 +60,13 @@ visualizer = dict(
     type='TextDetLocalVisualizer',
     name='visualizer',
     vis_backends=[dict(type='LocalVisBackend')])
-max_epochs = 200
+max_epochs = 50
 optim_wrapper = dict(
     type='OptimWrapper', optimizer=dict(type='Adam', lr=0.001))
 train_cfg = dict(type='EpochBasedTrainLoop', max_epochs=50, val_interval=20)
 val_cfg = dict(type='ValLoop')
 test_cfg = dict(type='TestLoop')
-param_scheduler = [dict(type='PolyLR', power=0.9, end=200)]
-thvotecount_textdet_train = dict(
-    type='OCRDataset',
-    data_root='data/det/vl+vc-textdet',
-    ann_file='textdet_train.json',
-    data_prefix=dict(img_path='imgs/'),
-    filter_cfg=dict(filter_empty_gt=True, min_size=32),
-    pipeline=[
-        dict(
-            type='LoadImageFromFile',
-            file_client_args=dict(backend='disk'),
-            color_type='color_ignore_orientation'),
-        dict(
-            type='LoadOCRAnnotations',
-            with_polygon=True,
-            with_bbox=True,
-            with_label=True),
-        dict(
-            type='TorchVisionWrapper',
-            op='ColorJitter',
-            brightness=0.12549019607843137,
-            saturation=0.5),
-        dict(type='FixInvalidPolygon'),
-        dict(type='ShortScaleAspectJitter', short_size=736, scale_divisor=32),
-        dict(type='RandomFlip', prob=0.5, direction='horizontal'),
-        dict(type='RandomRotate', max_angle=10),
-        dict(type='TextDetRandomCrop', target_size=(736, 736)),
-        dict(type='Pad', size=(736, 736)),
-        dict(
-            type='PackTextDetInputs',
-            meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
-    ])
-thvotecount_textdet_test = dict(
-    type='OCRDataset',
-    data_root='data/det/textdet-thvote',
-    ann_file='textdet_test.json',
-    data_prefix=dict(img_path='imgs/'),
-    test_mode=True,
-    pipeline=[
-        dict(
-            type='LoadImageFromFile',
-            file_client_args=dict(backend='disk'),
-            color_type='color_ignore_orientation'),
-        dict(type='Resize', scale=(2240, 2240), keep_ratio=True),
-        dict(
-            type='LoadOCRAnnotations',
-            with_polygon=True,
-            with_bbox=True,
-            with_label=True),
-        dict(
-            type='PackTextDetInputs',
-            meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
-    ])
+param_scheduler = [dict(type='PolyLR', power=0.9, end=50)]
 train_dataloader = dict(
     batch_size=10,
     num_workers=16,
@@ -258,7 +98,6 @@ train_dataloader = dict(
                 type='ShortScaleAspectJitter',
                 short_size=736,
                 scale_divisor=32),
-            dict(type='RandomFlip', prob=0.5, direction='horizontal'),
             dict(type='RandomRotate', max_angle=10),
             dict(type='TextDetRandomCrop', target_size=(736, 736)),
             dict(type='Pad', size=(736, 736)),