Skip to content

Commit

Permalink
init LOCE
Browse files Browse the repository at this point in the history
  • Loading branch information
@fcjian
fcjian committed Aug 18, 2021
1 parent fcb4601 commit d1b17bc
Show file tree
Hide file tree
Showing 4 changed files with 30 additions and 156 deletions.
1 change: 1 addition & 0 deletions configs/loce/loce_mask_rcnn_r50_fpn_normed_mask_mstrain_2x_lvis_v0.5.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,7 @@
total_epochs = 6

# custon hooks
# "InitializerHook" is defined in "mmdet/core/utils/initializer_hook.py"
custom_hooks = [
dict(type="InitializerHook")
]
Expand Down
1 change: 1 addition & 0 deletions configs/loce/loce_mask_rcnn_r50_fpn_normed_mask_mstrain_2x_lvis_v1.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -33,6 +33,7 @@
total_epochs = 6

# custon hooks
# "InitializerHook" is defined in "mmdet/core/utils/initializer_hook.py"
custom_hooks = [
dict(type="InitializerHook")
]
Expand Down
92 changes: 14 additions & 78 deletions configs/loce/mask_rcnn_r50_fpn_normed_mask_mstrain_2x_lvis_v0.5.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -134,11 +134,7 @@
multiscale_mode='value',
keep_ratio=True),
dict(type='RandomFlip', flip_ratio=0.5),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Normalize', **img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='DefaultFormatBundle'),
dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks'])
Expand All @@ -152,11 +148,7 @@
transforms=[
dict(type='Resize', keep_ratio=True),
dict(type='RandomFlip'),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Normalize', **img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='ImageToTensor', keys=['img']),
dict(type='Collect', keys=['img'])
Expand All @@ -166,76 +158,20 @@
samples_per_gpu=2,
workers_per_gpu=2,
train=dict(
type='LVISV05Dataset',
ann_file='data/lvis_v0.5/annotations/lvis_v0.5_train.json',
img_prefix='data/lvis_v0.5/train2017/',
pipeline=[
dict(type='LoadImageFromFile'),
dict(type='LoadAnnotations', with_bbox=True, with_mask=True),
dict(
type='Resize',
img_scale=[(1333, 640), (1333, 672), (1333, 704), (1333, 736),
(1333, 768), (1333, 800)],
multiscale_mode='value',
keep_ratio=True),
dict(type='RandomFlip', flip_ratio=0.5),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Pad', size_divisor=32),
dict(type='DefaultFormatBundle'),
dict(
type='Collect',
keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks'])
]),
type=dataset_type,
ann_file=data_root + 'annotations/lvis_v0.5_train.json',
img_prefix=data_root + 'train2017/',
pipeline=train_pipeline),
val=dict(
type='LVISV05Dataset',
ann_file='data/lvis_v0.5/annotations/lvis_v0.5_val.json',
img_prefix='data/lvis_v0.5/val2017/',
pipeline=[
dict(type='LoadImageFromFile'),
dict(
type='MultiScaleFlipAug',
img_scale=(1333, 800),
flip=False,
transforms=[
dict(type='Resize', keep_ratio=True),
dict(type='RandomFlip'),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Pad', size_divisor=32),
dict(type='ImageToTensor', keys=['img']),
dict(type='Collect', keys=['img'])
])
]),
type=dataset_type,
ann_file=data_root + 'annotations/lvis_v0.5_val.json',
img_prefix=data_root + 'val2017/',
pipeline=test_pipeline),
test=dict(
type='LVISV05Dataset',
ann_file='data/lvis_v0.5/annotations/lvis_v0.5_val.json',
img_prefix='data/lvis_v0.5/val2017/',
pipeline=[
dict(type='LoadImageFromFile'),
dict(
type='MultiScaleFlipAug',
img_scale=(1333, 800),
flip=False,
transforms=[
dict(type='Resize', keep_ratio=True),
dict(type='RandomFlip'),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Pad', size_divisor=32),
dict(type='ImageToTensor', keys=['img']),
dict(type='Collect', keys=['img'])
])
]))
type=dataset_type,
ann_file=data_root + 'annotations/lvis_v0.5_val.json',
img_prefix=data_root + 'val2017/',
pipeline=test_pipeline))
evaluation = dict(metric=['bbox', 'segm'], interval=12)
optimizer = dict(type='SGD', lr=0.02, momentum=0.9, weight_decay=0.0001)
optimizer_config = dict(grad_clip=None)
Expand Down
92 changes: 14 additions & 78 deletions configs/loce/mask_rcnn_r50_fpn_normed_mask_mstrain_2x_lvis_v1.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -134,11 +134,7 @@
multiscale_mode='value',
keep_ratio=True),
dict(type='RandomFlip', flip_ratio=0.5),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Normalize', **img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='DefaultFormatBundle'),
dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks'])
Expand All @@ -152,11 +148,7 @@
transforms=[
dict(type='Resize', keep_ratio=True),
dict(type='RandomFlip'),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Normalize', **img_norm_cfg),
dict(type='Pad', size_divisor=32),
dict(type='ImageToTensor', keys=['img']),
dict(type='Collect', keys=['img'])
Expand All @@ -166,76 +158,20 @@
samples_per_gpu=2,
workers_per_gpu=2,
train=dict(
type='LVISV1Dataset',
ann_file='data/lvis_v1/annotations/lvis_v1_train.json',
img_prefix='data/lvis_v1/',
pipeline=[
dict(type='LoadImageFromFile'),
dict(type='LoadAnnotations', with_bbox=True, with_mask=True),
dict(
type='Resize',
img_scale=[(1333, 640), (1333, 672), (1333, 704), (1333, 736),
(1333, 768), (1333, 800)],
multiscale_mode='value',
keep_ratio=True),
dict(type='RandomFlip', flip_ratio=0.5),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Pad', size_divisor=32),
dict(type='DefaultFormatBundle'),
dict(
type='Collect',
keys=['img', 'gt_bboxes', 'gt_labels', 'gt_masks'])
]),
type=dataset_type,
ann_file=data_root + 'annotations/lvis_v1_train.json',
img_prefix=data_root,
pipeline=train_pipeline),
val=dict(
type='LVISV1Dataset',
ann_file='data/lvis_v1/annotations/lvis_v1_val.json',
img_prefix='data/lvis_v1/',
pipeline=[
dict(type='LoadImageFromFile'),
dict(
type='MultiScaleFlipAug',
img_scale=(1333, 800),
flip=False,
transforms=[
dict(type='Resize', keep_ratio=True),
dict(type='RandomFlip'),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Pad', size_divisor=32),
dict(type='ImageToTensor', keys=['img']),
dict(type='Collect', keys=['img'])
])
]),
type=dataset_type,
ann_file=data_root + 'annotations/lvis_v1_val.json',
img_prefix=data_root,
pipeline=test_pipeline),
test=dict(
type='LVISV1Dataset',
ann_file='data/lvis_v1/annotations/lvis_v1_val.json',
img_prefix='data/lvis_v1/',
pipeline=[
dict(type='LoadImageFromFile'),
dict(
type='MultiScaleFlipAug',
img_scale=(1333, 800),
flip=False,
transforms=[
dict(type='Resize', keep_ratio=True),
dict(type='RandomFlip'),
dict(
type='Normalize',
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True),
dict(type='Pad', size_divisor=32),
dict(type='ImageToTensor', keys=['img']),
dict(type='Collect', keys=['img'])
])
]))
type=dataset_type,
ann_file=data_root + 'annotations/lvis_v1_val.json',
img_prefix=data_root,
pipeline=test_pipeline))
evaluation = dict(metric=['bbox', 'segm'], interval=12)
optimizer = dict(type='SGD', lr=0.02, momentum=0.9, weight_decay=0.0001)
optimizer_config = dict(grad_clip=None)
Expand Down

0 comments on commit d1b17bc

Please sign in to comment.