[Config] fix pan config (#1203)

configs/textdet/panet/_base_panet_r50_fpem_ffm.py

@@ -1,6 +1,5 @@
 model = dict(
     type='PANet',
-    pretrained='torchvision://resnet50',
     backbone=dict(
         type='mmdet.ResNet',
         depth=50,
@@ -9,13 +8,20 @@ model = dict(
         frozen_stages=1,
         norm_cfg=dict(type='BN', requires_grad=True),
         norm_eval=True,
-        style='caffe'),
+        style='caffe',
+        init_cfg=dict(type='Pretrained', checkpoint='torchvision://resnet50'),
+    ),
     neck=dict(type='FPEM_FFM', in_channels=[256, 512, 1024, 2048]),
-    bbox_head=dict(
+    det_head=dict(
         type='PANHead',
         in_channels=[128, 128, 128, 128],
-        out_channels=6,
-        module_loss=dict(type='PANModuleLoss', speedup_bbox_thr=32),
+        hidden_dim=128,
+        out_channel=6,
+        module_loss=dict(
+            type='PANModuleLoss',
+            loss_text=dict(type='MaskedSquareDiceLoss'),
+            loss_kernel=dict(type='MaskedSquareDiceLoss'),
+        ),
         postprocessor=dict(type='PANPostprocessor', text_repr_type='poly')),
     data_preprocessor=dict(
         type='TextDetDataPreprocessor',

configs/textdet/panet/panet_r18_fpem_ffm_600e_ctw1500.py

@@ -1,34 +1,78 @@
 _base_ = [
-    'panet_r18_fpem_ffm.py', '../../_base_/default_runtime.py',
-    '../../_base_/schedules/schedule_adam_600e.py',
+    '_base_panet_r18_fpem_ffm.py',
+    '../../_base_/default_runtime.py',
     '../../_base_/det_datasets/ctw1500.py',
-    '../../_base_/det_pipelines/panet_pipeline.py'
+    '../../_base_/schedules/schedule_adam_600e.py',
 ]
 
-model = {{_base_.model_poly}}
-
+# dataset settings
 train_list = {{_base_.train_list}}
 test_list = {{_base_.test_list}}
+file_client_args = dict(backend='disk')
+default_hooks = dict(checkpoint=dict(type='CheckpointHook', interval=20), )
 
-train_pipeline_ctw1500 = {{_base_.train_pipeline_ctw1500}}
-test_pipeline_ctw1500 = {{_base_.test_pipeline_ctw1500}}
+train_pipeline = [
+    dict(
+        type='LoadImageFromFile',
+        file_client_args=file_client_args,
+        color_type='color_ignore_orientation'),
+    dict(
+        type='LoadOCRAnnotations',
+        with_polygon=True,
+        with_bbox=True,
+        with_label=True,
+    ),
+    dict(type='ShortScaleAspectJitter', short_size=640, scale_divisor=32),
+    dict(type='RandomFlip', prob=0.5, direction='horizontal'),
+    dict(type='RandomRotate', max_angle=10),
+    dict(type='TextDetRandomCrop', target_size=(640, 640)),
+    dict(type='Pad', size=(640, 640)),
+    dict(
+        type='TorchVisionWrapper',
+        op='ColorJitter',
+        brightness=32.0 / 255,
+        saturation=0.5),
+    dict(
+        type='PackTextDetInputs',
+        meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
+]
 
-data = dict(
-    samples_per_gpu=2,
-    workers_per_gpu=2,
-    val_dataloader=dict(samples_per_gpu=1),
-    test_dataloader=dict(samples_per_gpu=1),
-    train=dict(
-        type='UniformConcatDataset',
-        datasets=train_list,
-        pipeline=train_pipeline_ctw1500),
-    val=dict(
-        type='UniformConcatDataset',
-        datasets=test_list,
-        pipeline=test_pipeline_ctw1500),
-    test=dict(
-        type='UniformConcatDataset',
-        datasets=test_list,
-        pipeline=test_pipeline_ctw1500))
+test_pipeline = [
+    dict(
+        type='LoadImageFromFile',
+        file_client_args=file_client_args,
+        color_type='color_ignore_orientation'),
+    # TODO Replace with mmcv.RescaleToShort when it's ready
+    dict(
+        type='ShortScaleAspectJitter',
+        short_size=640,
+        scale_divisor=1,
+        ratio_range=(1.0, 1.0),
+        aspect_ratio_range=(1.0, 1.0)),
+    dict(
+        type='PackTextDetInputs',
+        meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor',
+                   'instances'))
+]
+model = dict(det_head=dict(module_loss=dict(shrink_ratio=(1, 0.7))))
+train_dataloader = dict(
+    batch_size=64,
+    num_workers=8,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    dataset=dict(
+        type='ConcatDataset', datasets=train_list, pipeline=train_pipeline))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type='ConcatDataset', datasets=test_list, pipeline=test_pipeline))
+test_dataloader = val_dataloader
 
-evaluation = dict(interval=10, metric='hmean-iou')
+val_evaluator = dict(
+    type='HmeanIOUMetric', pred_score_thrs=dict(start=0.3, stop=1, step=0.05))
+test_evaluator = val_evaluator
+
+visualizer = dict(type='TextDetLocalVisualizer', name='visualizer')

configs/textdet/panet/panet_r18_fpem_ffm_600e_icdar2015.py

@@ -1,5 +1,5 @@
 _base_ = [
-    'panet_r18_fpem_ffm.py',
+    '_base_panet_r18_fpem_ffm.py',
     '../../_base_/default_runtime.py',
     '../../_base_/det_datasets/icdar2015.py',
     '../../_base_/schedules/schedule_adam_600e.py',
@@ -56,8 +56,8 @@ test_pipeline = [
 ]
 
 train_dataloader = dict(
-    batch_size=8,
-    num_workers=4,
+    batch_size=64,
+    num_workers=8,
     persistent_workers=True,
     sampler=dict(type='DefaultSampler', shuffle=True),
     dataset=dict(

configs/textdet/panet/panet_r50_fpem_ffm_600e_icdar2017.py

@@ -1,32 +1,78 @@
 _base_ = [
-    'panet_r50_fpem_ffm.py', '../../_base_/default_runtime.py',
-    '../../_base_/schedules/schedule_adam_600e.py',
+    '_base_panet_r50_fpem_ffm.py',
+    '../../_base_/default_runtime.py',
     '../../_base_/det_datasets/icdar2017.py',
-    '../../_base_/det_pipelines/panet_pipeline.py'
+    '../../_base_/schedules/schedule_adam_600e.py',
 ]
 
+# dataset settings
 train_list = {{_base_.train_list}}
 test_list = {{_base_.test_list}}
+file_client_args = dict(backend='disk')
+default_hooks = dict(checkpoint=dict(type='CheckpointHook', interval=20), )
 
-train_pipeline_icdar2017 = {{_base_.train_pipeline_icdar2017}}
-test_pipeline_icdar2017 = {{_base_.test_pipeline_icdar2017}}
+train_pipeline = [
+    dict(
+        type='LoadImageFromFile',
+        file_client_args=file_client_args,
+        color_type='color_ignore_orientation'),
+    dict(
+        type='LoadOCRAnnotations',
+        with_polygon=True,
+        with_bbox=True,
+        with_label=True,
+    ),
+    dict(type='ShortScaleAspectJitter', short_size=800, scale_divisor=32),
+    dict(type='RandomFlip', prob=0.5, direction='horizontal'),
+    dict(type='RandomRotate', max_angle=10),
+    dict(type='TextDetRandomCrop', target_size=(800, 800)),
+    dict(type='Pad', size=(800, 800)),
+    dict(
+        type='TorchVisionWrapper',
+        op='ColorJitter',
+        brightness=32.0 / 255,
+        saturation=0.5),
+    dict(
+        type='PackTextDetInputs',
+        meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor'))
+]
 
-data = dict(
-    samples_per_gpu=4,
-    workers_per_gpu=4,
-    val_dataloader=dict(samples_per_gpu=1),
-    test_dataloader=dict(samples_per_gpu=1),
-    train=dict(
-        type='UniformConcatDataset',
-        datasets=train_list,
-        pipeline=train_pipeline_icdar2017),
-    val=dict(
-        type='UniformConcatDataset',
-        datasets=test_list,
-        pipeline=test_pipeline_icdar2017),
-    test=dict(
-        type='UniformConcatDataset',
-        datasets=test_list,
-        pipeline=test_pipeline_icdar2017))
+test_pipeline = [
+    dict(
+        type='LoadImageFromFile',
+        file_client_args=file_client_args,
+        color_type='color_ignore_orientation'),
+    # TODO Replace with mmcv.RescaleToShort when it's ready
+    dict(
+        type='ShortScaleAspectJitter',
+        short_size=800,
+        scale_divisor=1,
+        ratio_range=(1.0, 1.0),
+        aspect_ratio_range=(1.0, 1.0)),
+    dict(
+        type='PackTextDetInputs',
+        meta_keys=('img_path', 'ori_shape', 'img_shape', 'scale_factor',
+                   'instances'))
+]
 
-evaluation = dict(interval=10, metric='hmean-iou')
+train_dataloader = dict(
+    batch_size=64,
+    num_workers=8,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=True),
+    dataset=dict(
+        type='ConcatDataset', datasets=train_list, pipeline=train_pipeline))
+val_dataloader = dict(
+    batch_size=1,
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(type='DefaultSampler', shuffle=False),
+    dataset=dict(
+        type='ConcatDataset', datasets=test_list, pipeline=test_pipeline))
+test_dataloader = val_dataloader
+
+val_evaluator = dict(
+    type='HmeanIOUMetric', pred_score_thrs=dict(start=0.3, stop=1, step=0.05))
+test_evaluator = val_evaluator
+
+visualizer = dict(type='TextDetLocalVisualizer', name='visualizer')