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

[Enhancement] Add codespell pre-commit hook and fix typos (#920) 

* add codespell pre-commit hook and fix typos

* Update mmseg/models/decode_heads/dpt_head.py

* Update mmseg/models/backbones/vit.py

* Update mmseg/models/backbones/vit.py

* fix typos

* skip formating typo

* deprecate formating

* skip ipynb

* unstage ipynb changes

* unstage ipynb changes

* fix typos in ipynb

* unstage ipynb changes
pull/1801/head
Junjun2016 2021-10-13 21:21:17 +08:00 committed by GitHub
parent 279b7e5d25
commit 67f1420472
23 changed files with 352 additions and 333 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
.github/ISSUE_TEMPLATE/config.yml vendored
View File

@ -3,4 +3,4 @@ blank_issues_enabled: false
contact_links:
- name: MMSegmentation Documentation
url: https://mmsegmentation.readthedocs.io
about: Check the docs and FAQ to see if you question is already anwsered.
about: Check the docs and FAQ to see if you question is already answered.

2
.github/ISSUE_TEMPLATE/error-report.md vendored
View File

@ -30,7 +30,7 @@ A clear and concise description of what the bug is.
**Environment**
1. Please run `python mmseg/utils/collect_env.py` to collect necessary environment infomation and paste it here.
1. Please run `python mmseg/utils/collect_env.py` to collect necessary environment information and paste it here.
2. You may add addition that may be helpful for locating the problem, such as
- How you installed PyTorch [e.g., pip, conda, source]
- Other environment variables that may be related (such as `$PATH`, `$LD_LIBRARY_PATH`, `$PYTHONPATH`, etc.)

4
.pre-commit-config.yaml
View File

@ -34,6 +34,10 @@ repos:
hooks:
- id: markdownlint
args: ["-r", "~MD002,~MD013,~MD029,~MD033,~MD034,~MD036"]
- repo: https://github.com/codespell-project/codespell
rev: v2.1.0
hooks:
- id: codespell
- repo: https://github.com/myint/docformatter
rev: v1.3.1
hooks:

2
README.md
View File

@ -124,7 +124,7 @@ Please refer to [get_started.md](docs/get_started.md#installation) for installat
Please see [train.md](docs/train.md) and [inference.md](docs/inference.md) for the basic usage of MMSegmentation.
There are also tutorials for [customizing dataset](docs/tutorials/customize_datasets.md), [designing data pipeline](docs/tutorials/data_pipeline.md), [customizing modules](docs/tutorials/customize_models.md), and [customizing runtime](docs/tutorials/customize_runtime.md).
We also provide many [training tricks](docs/tutorials/training_tricks.md) for better training and [usefule tools](docs/useful_tools.md) for deployment.
We also provide many [training tricks](docs/tutorials/training_tricks.md) for better training and [useful tools](docs/useful_tools.md) for deployment.
A Colab tutorial is also provided. You may preview the notebook [here](demo/MMSegmentation_Tutorial.ipynb) or directly [run](https://colab.research.google.com/github/open-mmlab/mmsegmentation/blob/master/demo/MMSegmentation_Tutorial.ipynb) on Colab.

8
docs/tutorials/config.md
View File

@ -67,7 +67,7 @@ model = dict(
channels=512, # The intermediate channels of decode head.
pool_scales=(1, 2, 3, 6), # The avg pooling scales of PSPHead. Please refer to paper for details.
dropout_ratio=0.1, # The dropout ratio before final classification layer.
num_classes=19, # Number of segmentation classs. Usually 19 for cityscapes, 21 for VOC, 150 for ADE20k.
num_classes=19, # Number of segmentation class. Usually 19 for cityscapes, 21 for VOC, 150 for ADE20k.
norm_cfg=dict(type='SyncBN', requires_grad=True), # The configuration of norm layer.
align_corners=False, # The align_corners argument for resize in decoding.
loss_decode=dict( # Config of loss function for the decode_head.
@ -82,7 +82,7 @@ model = dict(
num_convs=1, # Number of convs in FCNHead. It is usually 1 in auxiliary head.
concat_input=False, # Whether concat output of convs with input before classification layer.
dropout_ratio=0.1, # The dropout ratio before final classification layer.
num_classes=19, # Number of segmentation classs. Usually 19 for cityscapes, 21 for VOC, 150 for ADE20k.
num_classes=19, # Number of segmentation class. Usually 19 for cityscapes, 21 for VOC, 150 for ADE20k.
norm_cfg=dict(type='SyncBN', requires_grad=True), # The configuration of norm layer.
align_corners=False, # The align_corners argument for resize in decoding.
loss_decode=dict( # Config of loss function for the decode_head.
@ -132,7 +132,7 @@ test_pipeline = [
flip=False, # Whether to flip images during testing
transforms=[
dict(type='Resize', # Use resize augmentation
keep_ratio=True), # Whether to keep the ratio between height and width, the img_scale set here will be supressed by the img_scale set above.
keep_ratio=True), # Whether to keep the ratio between height and width, the img_scale set here will be suppressed by the img_scale set above.
dict(type='RandomFlip'), # Thought RandomFlip is added in pipeline, it is not used when flip=False
dict(
type='Normalize', # Normalization config, the values are from img_norm_cfg
@ -245,7 +245,7 @@ runner = dict(
checkpoint_config = dict( # Config to set the checkpoint hook, Refer to https://github.com/open-mmlab/mmcv/blob/master/mmcv/runner/hooks/checkpoint.py for implementation.
by_epoch=False, # Whether count by epoch or not.
interval=4000) # The save interval.
evaluation = dict( # The config to build the evaluation hook. Please refer to mmseg/core/evaulation/eval_hook.py for details.
evaluation = dict( # The config to build the evaluation hook. Please refer to mmseg/core/evaluation/eval_hook.py for details.
interval=4000, # The interval of evaluation.
metric='mIoU') # The evaluation metric.

4
docs/tutorials/customize_runtime.md
View File

@ -113,7 +113,7 @@ Tricks not implemented by the optimizer should be implemented through optimizer
_delete_=True, grad_clip=dict(max_norm=35, norm_type=2))
```
If your config inherits the base config which already sets the `optimizer_config`, you might need `_delete_=True` to overide the unnecessary settings. See the [config documenetation](https://mmsegmentation.readthedocs.io/en/latest/config.html) for more details.
If your config inherits the base config which already sets the `optimizer_config`, you might need `_delete_=True` to override the unnecessary settings. See the [config documentation](https://mmsegmentation.readthedocs.io/en/latest/config.html) for more details.
- __Use momentum schedule to accelerate model convergence__:
We support momentum scheduler to modify model's momentum according to learning rate, which could make the model converge in a faster way.
@ -198,7 +198,7 @@ custom_hooks = [
### Modify default runtime hooks
There are some common hooks that are not registerd through `custom_hooks`, they are
There are some common hooks that are not registered through `custom_hooks`, they are
- log_config
- checkpoint_config

2
docs_zh-CN/tutorials/config.md
View File

@ -241,7 +241,7 @@ runner = dict(
checkpoint_config = dict( # 设置检查点钩子 (checkpoint hook) 的配置文件。执行时请参考 https://github.com/open-mmlab/mmcv/blob/master/mmcv/runner/hooks/checkpoint.py。
by_epoch=False, # 是否按照每个 epoch 去算 runner。
interval=4000) # 保存的间隔
evaluation = dict( # 构建评估钩 (evaluation hook) 的配置文件。细节请参考 mmseg/core/evaulation/eval_hook.py。
evaluation = dict( # 构建评估钩 (evaluation hook) 的配置文件。细节请参考 mmseg/core/evaluation/eval_hook.py。
interval=4000, # 评估的间歇点
metric='mIoU') # 评估的指标

14
mmseg/core/evaluation/metrics.py
View File

@ -7,7 +7,7 @@ import torch
def f_score(precision, recall, beta=1):
"""calcuate the f-score value.
"""calculate the f-score value.
Args:
precision (float | torch.Tensor): The precision value.
@ -40,7 +40,7 @@ def intersect_and_union(pred_label,
ignore_index (int): Index that will be ignored in evaluation.
label_map (dict): Mapping old labels to new labels. The parameter will
work only when label is str. Default: dict().
reduce_zero_label (bool): Wether ignore zero label. The parameter will
reduce_zero_label (bool): Whether ignore zero label. The parameter will
work only when label is str. Default: False.
Returns:
@ -102,7 +102,7 @@ def total_intersect_and_union(results,
num_classes (int): Number of categories.
ignore_index (int): Index that will be ignored in evaluation.
label_map (dict): Mapping old labels to new labels. Default: dict().
reduce_zero_label (bool): Wether ignore zero label. Default: False.
reduce_zero_label (bool): Whether ignore zero label. Default: False.
Returns:
ndarray: The intersection of prediction and ground truth histogram
@ -148,7 +148,7 @@ def mean_iou(results,
nan_to_num (int, optional): If specified, NaN values will be replaced
by the numbers defined by the user. Default: None.
label_map (dict): Mapping old labels to new labels. Default: dict().
reduce_zero_label (bool): Wether ignore zero label. Default: False.
reduce_zero_label (bool): Whether ignore zero label. Default: False.
Returns:
dict[str, float | ndarray]:
@ -187,7 +187,7 @@ def mean_dice(results,
nan_to_num (int, optional): If specified, NaN values will be replaced
by the numbers defined by the user. Default: None.
label_map (dict): Mapping old labels to new labels. Default: dict().
reduce_zero_label (bool): Wether ignore zero label. Default: False.
reduce_zero_label (bool): Whether ignore zero label. Default: False.
Returns:
dict[str, float | ndarray]: Default metrics.
@ -228,7 +228,7 @@ def mean_fscore(results,
nan_to_num (int, optional): If specified, NaN values will be replaced
by the numbers defined by the user. Default: None.
label_map (dict): Mapping old labels to new labels. Default: dict().
reduce_zero_label (bool): Wether ignore zero label. Default: False.
reduce_zero_label (bool): Whether ignore zero label. Default: False.
beta (int): Determines the weight of recall in the combined score.
Default: False.
@ -274,7 +274,7 @@ def eval_metrics(results,
nan_to_num (int, optional): If specified, NaN values will be replaced
by the numbers defined by the user. Default: None.
label_map (dict): Mapping old labels to new labels. Default: dict().
reduce_zero_label (bool): Wether ignore zero label. Default: False.
reduce_zero_label (bool): Whether ignore zero label. Default: False.
Returns:
float: Overall accuracy on all images.
ndarray: Per category accuracy, shape (num_classes, ).

4
mmseg/datasets/pipelines/__init__.py
View File

@ -1,7 +1,7 @@
# Copyright (c) OpenMMLab. All rights reserved.
from .compose import Compose
from .formating import (Collect, ImageToTensor, ToDataContainer, ToTensor,
Transpose, to_tensor)
from .formatting import (Collect, ImageToTensor, ToDataContainer, ToTensor,
Transpose, to_tensor)
from .loading import LoadAnnotations, LoadImageFromFile
from .test_time_aug import MultiScaleFlipAug
from .transforms import (CLAHE, AdjustGamma, Normalize, Pad,

292
mmseg/datasets/pipelines/formating.py
View File

@ -1,289 +1,9 @@
# Copyright (c) OpenMMLab. All rights reserved.
from collections.abc import Sequence
# flake8: noqa
import warnings
import mmcv
import numpy as np
import torch
from mmcv.parallel import DataContainer as DC
from .formatting import *
from ..builder import PIPELINES
def to_tensor(data):
"""Convert objects of various python types to :obj:`torch.Tensor`.
Supported types are: :class:`numpy.ndarray`, :class:`torch.Tensor`,
:class:`Sequence`, :class:`int` and :class:`float`.
Args:
data (torch.Tensor | numpy.ndarray | Sequence | int | float): Data to
be converted.
"""
if isinstance(data, torch.Tensor):
return data
elif isinstance(data, np.ndarray):
return torch.from_numpy(data)
elif isinstance(data, Sequence) and not mmcv.is_str(data):
return torch.tensor(data)
elif isinstance(data, int):
return torch.LongTensor([data])
elif isinstance(data, float):
return torch.FloatTensor([data])
else:
raise TypeError(f'type {type(data)} cannot be converted to tensor.')
@PIPELINES.register_module()
class ToTensor(object):
"""Convert some results to :obj:`torch.Tensor` by given keys.
Args:
keys (Sequence[str]): Keys that need to be converted to Tensor.
"""
def __init__(self, keys):
self.keys = keys
def __call__(self, results):
"""Call function to convert data in results to :obj:`torch.Tensor`.
Args:
results (dict): Result dict contains the data to convert.
Returns:
dict: The result dict contains the data converted
to :obj:`torch.Tensor`.
"""
for key in self.keys:
results[key] = to_tensor(results[key])
return results
def __repr__(self):
return self.__class__.__name__ + f'(keys={self.keys})'
@PIPELINES.register_module()
class ImageToTensor(object):
"""Convert image to :obj:`torch.Tensor` by given keys.
The dimension order of input image is (H, W, C). The pipeline will convert
it to (C, H, W). If only 2 dimension (H, W) is given, the output would be
(1, H, W).
Args:
keys (Sequence[str]): Key of images to be converted to Tensor.
"""
def __init__(self, keys):
self.keys = keys
def __call__(self, results):
"""Call function to convert image in results to :obj:`torch.Tensor` and
transpose the channel order.
Args:
results (dict): Result dict contains the image data to convert.
Returns:
dict: The result dict contains the image converted
to :obj:`torch.Tensor` and transposed to (C, H, W) order.
"""
for key in self.keys:
img = results[key]
if len(img.shape) < 3:
img = np.expand_dims(img, -1)
results[key] = to_tensor(img.transpose(2, 0, 1))
return results
def __repr__(self):
return self.__class__.__name__ + f'(keys={self.keys})'
@PIPELINES.register_module()
class Transpose(object):
"""Transpose some results by given keys.
Args:
keys (Sequence[str]): Keys of results to be transposed.
order (Sequence[int]): Order of transpose.
"""
def __init__(self, keys, order):
self.keys = keys
self.order = order
def __call__(self, results):
"""Call function to convert image in results to :obj:`torch.Tensor` and
transpose the channel order.
Args:
results (dict): Result dict contains the image data to convert.
Returns:
dict: The result dict contains the image converted
to :obj:`torch.Tensor` and transposed to (C, H, W) order.
"""
for key in self.keys:
results[key] = results[key].transpose(self.order)
return results
def __repr__(self):
return self.__class__.__name__ + \
f'(keys={self.keys}, order={self.order})'
@PIPELINES.register_module()
class ToDataContainer(object):
"""Convert results to :obj:`mmcv.DataContainer` by given fields.
Args:
fields (Sequence[dict]): Each field is a dict like
``dict(key='xxx', **kwargs)``. The ``key`` in result will
be converted to :obj:`mmcv.DataContainer` with ``**kwargs``.
Default: ``(dict(key='img', stack=True),
dict(key='gt_semantic_seg'))``.
"""
def __init__(self,
fields=(dict(key='img',
stack=True), dict(key='gt_semantic_seg'))):
self.fields = fields
def __call__(self, results):
"""Call function to convert data in results to
:obj:`mmcv.DataContainer`.
Args:
results (dict): Result dict contains the data to convert.
Returns:
dict: The result dict contains the data converted to
:obj:`mmcv.DataContainer`.
"""
for field in self.fields:
field = field.copy()
key = field.pop('key')
results[key] = DC(results[key], **field)
return results
def __repr__(self):
return self.__class__.__name__ + f'(fields={self.fields})'
@PIPELINES.register_module()
class DefaultFormatBundle(object):
"""Default formatting bundle.
It simplifies the pipeline of formatting common fields, including "img"
and "gt_semantic_seg". These fields are formatted as follows.
- img: (1)transpose, (2)to tensor, (3)to DataContainer (stack=True)
- gt_semantic_seg: (1)unsqueeze dim-0 (2)to tensor,
(3)to DataContainer (stack=True)
"""
def __call__(self, results):
"""Call function to transform and format common fields in results.
Args:
results (dict): Result dict contains the data to convert.
Returns:
dict: The result dict contains the data that is formatted with
default bundle.
"""
if 'img' in results:
img = results['img']
if len(img.shape) < 3:
img = np.expand_dims(img, -1)
img = np.ascontiguousarray(img.transpose(2, 0, 1))
results['img'] = DC(to_tensor(img), stack=True)
if 'gt_semantic_seg' in results:
# convert to long
results['gt_semantic_seg'] = DC(
to_tensor(results['gt_semantic_seg'][None,
...].astype(np.int64)),
stack=True)
return results
def __repr__(self):
return self.__class__.__name__
@PIPELINES.register_module()
class Collect(object):
"""Collect data from the loader relevant to the specific task.
This is usually the last stage of the data loader pipeline. Typically keys
is set to some subset of "img", "gt_semantic_seg".
The "img_meta" item is always populated. The contents of the "img_meta"
dictionary depends on "meta_keys". By default this includes:
- "img_shape": shape of the image input to the network as a tuple
(h, w, c). Note that images may be zero padded on the bottom/right
if the batch tensor is larger than this shape.
- "scale_factor": a float indicating the preprocessing scale
- "flip": a boolean indicating if image flip transform was used
- "filename": path to the image file
- "ori_shape": original shape of the image as a tuple (h, w, c)
- "pad_shape": image shape after padding
- "img_norm_cfg": a dict of normalization information:
- mean - per channel mean subtraction
- std - per channel std divisor
- to_rgb - bool indicating if bgr was converted to rgb
Args:
keys (Sequence[str]): Keys of results to be collected in ``data``.
meta_keys (Sequence[str], optional): Meta keys to be converted to
``mmcv.DataContainer`` and collected in ``data[img_metas]``.
Default: (``filename``, ``ori_filename``, ``ori_shape``,
``img_shape``, ``pad_shape``, ``scale_factor``, ``flip``,
``flip_direction``, ``img_norm_cfg``)
"""
def __init__(self,
keys,
meta_keys=('filename', 'ori_filename', 'ori_shape',
'img_shape', 'pad_shape', 'scale_factor', 'flip',
'flip_direction', 'img_norm_cfg')):
self.keys = keys
self.meta_keys = meta_keys
def __call__(self, results):
"""Call function to collect keys in results. The keys in ``meta_keys``
will be converted to :obj:mmcv.DataContainer.
Args:
results (dict): Result dict contains the data to collect.
Returns:
dict: The result dict contains the following keys
- keys in``self.keys``
- ``img_metas``
"""
data = {}
img_meta = {}
for key in self.meta_keys:
img_meta[key] = results[key]
data['img_metas'] = DC(img_meta, cpu_only=True)
for key in self.keys:
data[key] = results[key]
return data
def __repr__(self):
return self.__class__.__name__ + \
f'(keys={self.keys}, meta_keys={self.meta_keys})'
warnings.warn('DeprecationWarning: mmseg.datasets.pipelines.formating will be '
'deprecated in 2021, please replace it with '
'mmseg.datasets.pipelines.formatting.')

289
mmseg/datasets/pipelines/formatting.py 100644
View File

@ -0,0 +1,289 @@
# Copyright (c) OpenMMLab. All rights reserved.
from collections.abc import Sequence
import mmcv
import numpy as np
import torch
from mmcv.parallel import DataContainer as DC
from ..builder import PIPELINES
def to_tensor(data):
"""Convert objects of various python types to :obj:`torch.Tensor`.
Supported types are: :class:`numpy.ndarray`, :class:`torch.Tensor`,
:class:`Sequence`, :class:`int` and :class:`float`.
Args:
data (torch.Tensor | numpy.ndarray | Sequence | int | float): Data to
be converted.
"""
if isinstance(data, torch.Tensor):
return data
elif isinstance(data, np.ndarray):
return torch.from_numpy(data)
elif isinstance(data, Sequence) and not mmcv.is_str(data):
return torch.tensor(data)
elif isinstance(data, int):
return torch.LongTensor([data])
elif isinstance(data, float):
return torch.FloatTensor([data])
else:
raise TypeError(f'type {type(data)} cannot be converted to tensor.')
@PIPELINES.register_module()
class ToTensor(object):
"""Convert some results to :obj:`torch.Tensor` by given keys.
Args:
keys (Sequence[str]): Keys that need to be converted to Tensor.
"""
def __init__(self, keys):
self.keys = keys
def __call__(self, results):
"""Call function to convert data in results to :obj:`torch.Tensor`.
Args:
results (dict): Result dict contains the data to convert.
Returns:
dict: The result dict contains the data converted
to :obj:`torch.Tensor`.
"""
for key in self.keys:
results[key] = to_tensor(results[key])
return results
def __repr__(self):
return self.__class__.__name__ + f'(keys={self.keys})'
@PIPELINES.register_module()
class ImageToTensor(object):
"""Convert image to :obj:`torch.Tensor` by given keys.
The dimension order of input image is (H, W, C). The pipeline will convert
it to (C, H, W). If only 2 dimension (H, W) is given, the output would be
(1, H, W).
Args:
keys (Sequence[str]): Key of images to be converted to Tensor.
"""
def __init__(self, keys):
self.keys = keys
def __call__(self, results):
"""Call function to convert image in results to :obj:`torch.Tensor` and
transpose the channel order.
Args:
results (dict): Result dict contains the image data to convert.
Returns:
dict: The result dict contains the image converted
to :obj:`torch.Tensor` and transposed to (C, H, W) order.
"""
for key in self.keys:
img = results[key]
if len(img.shape) < 3:
img = np.expand_dims(img, -1)
results[key] = to_tensor(img.transpose(2, 0, 1))
return results
def __repr__(self):
return self.__class__.__name__ + f'(keys={self.keys})'
@PIPELINES.register_module()
class Transpose(object):
"""Transpose some results by given keys.
Args:
keys (Sequence[str]): Keys of results to be transposed.
order (Sequence[int]): Order of transpose.
"""
def __init__(self, keys, order):
self.keys = keys
self.order = order
def __call__(self, results):
"""Call function to convert image in results to :obj:`torch.Tensor` and
transpose the channel order.
Args:
results (dict): Result dict contains the image data to convert.
Returns:
dict: The result dict contains the image converted
to :obj:`torch.Tensor` and transposed to (C, H, W) order.
"""
for key in self.keys:
results[key] = results[key].transpose(self.order)
return results
def __repr__(self):
return self.__class__.__name__ + \
f'(keys={self.keys}, order={self.order})'
@PIPELINES.register_module()
class ToDataContainer(object):
"""Convert results to :obj:`mmcv.DataContainer` by given fields.
Args:
fields (Sequence[dict]): Each field is a dict like
``dict(key='xxx', **kwargs)``. The ``key`` in result will
be converted to :obj:`mmcv.DataContainer` with ``**kwargs``.
Default: ``(dict(key='img', stack=True),
dict(key='gt_semantic_seg'))``.
"""
def __init__(self,
fields=(dict(key='img',
stack=True), dict(key='gt_semantic_seg'))):
self.fields = fields
def __call__(self, results):
"""Call function to convert data in results to
:obj:`mmcv.DataContainer`.
Args:
results (dict): Result dict contains the data to convert.
Returns:
dict: The result dict contains the data converted to
:obj:`mmcv.DataContainer`.
"""
for field in self.fields:
field = field.copy()
key = field.pop('key')
results[key] = DC(results[key], **field)
return results
def __repr__(self):
return self.__class__.__name__ + f'(fields={self.fields})'
@PIPELINES.register_module()
class DefaultFormatBundle(object):
"""Default formatting bundle.
It simplifies the pipeline of formatting common fields, including "img"
and "gt_semantic_seg". These fields are formatted as follows.
- img: (1)transpose, (2)to tensor, (3)to DataContainer (stack=True)
- gt_semantic_seg: (1)unsqueeze dim-0 (2)to tensor,
(3)to DataContainer (stack=True)
"""
def __call__(self, results):
"""Call function to transform and format common fields in results.
Args:
results (dict): Result dict contains the data to convert.
Returns:
dict: The result dict contains the data that is formatted with
default bundle.
"""
if 'img' in results:
img = results['img']
if len(img.shape) < 3:
img = np.expand_dims(img, -1)
img = np.ascontiguousarray(img.transpose(2, 0, 1))
results['img'] = DC(to_tensor(img), stack=True)
if 'gt_semantic_seg' in results:
# convert to long
results['gt_semantic_seg'] = DC(
to_tensor(results['gt_semantic_seg'][None,
...].astype(np.int64)),
stack=True)
return results
def __repr__(self):
return self.__class__.__name__
@PIPELINES.register_module()
class Collect(object):
"""Collect data from the loader relevant to the specific task.
This is usually the last stage of the data loader pipeline. Typically keys
is set to some subset of "img", "gt_semantic_seg".
The "img_meta" item is always populated. The contents of the "img_meta"
dictionary depends on "meta_keys". By default this includes:
- "img_shape": shape of the image input to the network as a tuple
(h, w, c). Note that images may be zero padded on the bottom/right
if the batch tensor is larger than this shape.
- "scale_factor": a float indicating the preprocessing scale
- "flip": a boolean indicating if image flip transform was used
- "filename": path to the image file
- "ori_shape": original shape of the image as a tuple (h, w, c)
- "pad_shape": image shape after padding
- "img_norm_cfg": a dict of normalization information:
- mean - per channel mean subtraction
- std - per channel std divisor
- to_rgb - bool indicating if bgr was converted to rgb
Args:
keys (Sequence[str]): Keys of results to be collected in ``data``.
meta_keys (Sequence[str], optional): Meta keys to be converted to
``mmcv.DataContainer`` and collected in ``data[img_metas]``.
Default: (``filename``, ``ori_filename``, ``ori_shape``,
``img_shape``, ``pad_shape``, ``scale_factor``, ``flip``,
``flip_direction``, ``img_norm_cfg``)
"""
def __init__(self,
keys,
meta_keys=('filename', 'ori_filename', 'ori_shape',
'img_shape', 'pad_shape', 'scale_factor', 'flip',
'flip_direction', 'img_norm_cfg')):
self.keys = keys
self.meta_keys = meta_keys
def __call__(self, results):
"""Call function to collect keys in results. The keys in ``meta_keys``
will be converted to :obj:mmcv.DataContainer.
Args:
results (dict): Result dict contains the data to collect.
Returns:
dict: The result dict contains the following keys
- keys in``self.keys``
- ``img_metas``
"""
data = {}
img_meta = {}
for key in self.meta_keys:
img_meta[key] = results[key]
data['img_metas'] = DC(img_meta, cpu_only=True)
for key in self.keys:
data[key] = results[key]
return data
def __repr__(self):
return self.__class__.__name__ + \
f'(keys={self.keys}, meta_keys={self.meta_keys})'

2
mmseg/models/backbones/bisenetv2.py
View File

@ -194,7 +194,7 @@ class GELayer(BaseModule):
init_cfg (dict or list[dict], optional): Initialization config dict.
Default: None.
Returns:
x (torch.Tensor): Intermidiate feature map in
x (torch.Tensor): Intermediate feature map in
Semantic Branch.
"""

4
mmseg/models/backbones/mit.py
View File

@ -186,7 +186,7 @@ class TransformerEncoderLayer(BaseModule):
qkv_bias (bool): enable bias for qkv if True.
Default: True.
act_cfg (dict): The activation config for FFNs.
Defalut: dict(type='GELU').
Default: dict(type='GELU').
norm_cfg (dict): Config dict for normalization layer.
Default: dict(type='LN').
batch_first (bool): Key, Query and Value are shape of
@ -277,7 +277,7 @@ class MixVisionTransformer(BaseModule):
norm_cfg (dict): Config dict for normalization layer.
Default: dict(type='LN')
act_cfg (dict): The activation config for FFNs.
Defalut: dict(type='GELU').
Default: dict(type='GELU').
pretrained (str, optional): model pretrained path. Default: None.
init_cfg (dict or list[dict], optional): Initialization config dict.
Default: None.

4
mmseg/models/backbones/vit.py
View File

@ -33,7 +33,7 @@ class TransformerEncoderLayer(BaseModule):
Default: 2.
qkv_bias (bool): enable bias for qkv if True. Default: True
act_cfg (dict): The activation config for FFNs.
Defalut: dict(type='GELU').
Default: dict(type='GELU').
norm_cfg (dict): Config dict for normalization layer.
Default: dict(type='LN').
batch_first (bool): Key, Query and Value are shape of
@ -126,7 +126,7 @@ class VisionTransformer(BaseModule):
norm_cfg (dict): Config dict for normalization layer.
Default: dict(type='LN')
act_cfg (dict): The activation config for FFNs.
Defalut: dict(type='GELU').
Default: dict(type='GELU').
patch_norm (bool): Whether to add a norm in PatchEmbed Block.
Default: False.
final_norm (bool): Whether to add a additional layer to normalize

2
mmseg/models/decode_heads/dpt_head.py
View File

@ -227,7 +227,7 @@ class DPTHead(BaseDecodeHead):
expand_channels (bool): Whether expand the channels in post process
block. Default: False.
act_cfg (dict): The activation config for residual conv unit.
Defalut dict(type='ReLU').
Default dict(type='ReLU').
norm_cfg (dict): Config dict for normalization layer.
Default: dict(type='BN').
"""

2
mmseg/models/losses/utils.py
View File

@ -51,7 +51,7 @@ def weight_reduce_loss(loss, weight=None, reduction='mean', avg_factor=None):
loss (Tensor): Element-wise loss.
weight (Tensor): Element-wise weights.
reduction (str): Same as built-in losses of PyTorch.
avg_factor (float): Avarage factor when computing the mean of losses.
avg_factor (float): Average factor when computing the mean of losses.
Returns:
Tensor: Processed loss values.

8
mmseg/models/utils/shape_convert.py
View File

@ -3,11 +3,11 @@ def nlc_to_nchw(x, hw_shape):
"""Convert [N, L, C] shape tensor to [N, C, H, W] shape tensor.
Args:
x (Tensor): The input tensor of shape [N, L, C] before convertion.
x (Tensor): The input tensor of shape [N, L, C] before conversion.
hw_shape (Sequence[int]): The height and width of output feature map.
Returns:
Tensor: The output tensor of shape [N, C, H, W] after convertion.
Tensor: The output tensor of shape [N, C, H, W] after conversion.
"""
H, W = hw_shape
assert len(x.shape) == 3
@ -20,10 +20,10 @@ def nchw_to_nlc(x):
"""Flatten [N, C, H, W] shape tensor to [N, L, C] shape tensor.
Args:
x (Tensor): The input tensor of shape [N, C, H, W] before convertion.
x (Tensor): The input tensor of shape [N, C, H, W] before conversion.
Returns:
Tensor: The output tensor of shape [N, L, C] after convertion.
Tensor: The output tensor of shape [N, L, C] after conversion.
"""
assert len(x.shape) == 4
return x.flatten(2).transpose(1, 2).contiguous()

6
setup.cfg
View File

@ -11,3 +11,9 @@ known_first_party = mmseg
known_third_party = PIL,cityscapesscripts,cv2,detail,matplotlib,mmcv,numpy,onnxruntime,packaging,prettytable,pytest,pytorch_sphinx_theme,requests,scipy,seaborn,torch,ts
no_lines_before = STDLIB,LOCALFOLDER
default_section = THIRDPARTY
[codespell]
skip = *.po,*.ts,*.ipynb
count =
quiet-level = 3
ignore-words-list = formating,sur,hist

12
tests/test_models/test_backbones/test_unet.py
View File

@ -425,7 +425,7 @@ def test_unet():
unet(x)
with pytest.raises(AssertionError):
# Check if num_stages matchs strides, len(strides)=num_stages
# Check if num_stages matches strides, len(strides)=num_stages
unet = UNet(
in_channels=3,
base_channels=4,
@ -440,7 +440,7 @@ def test_unet():
unet(x)
with pytest.raises(AssertionError):
# Check if num_stages matchs strides, len(enc_num_convs)=num_stages
# Check if num_stages matches strides, len(enc_num_convs)=num_stages
unet = UNet(
in_channels=3,
base_channels=4,
@ -455,7 +455,7 @@ def test_unet():
unet(x)
with pytest.raises(AssertionError):
# Check if num_stages matchs strides, len(dec_num_convs)=num_stages-1
# Check if num_stages matches strides, len(dec_num_convs)=num_stages-1
unet = UNet(
in_channels=3,
base_channels=4,
@ -470,7 +470,7 @@ def test_unet():
unet(x)
with pytest.raises(AssertionError):
# Check if num_stages matchs strides, len(downsamples)=num_stages-1
# Check if num_stages matches strides, len(downsamples)=num_stages-1
unet = UNet(
in_channels=3,
base_channels=4,
@ -485,7 +485,7 @@ def test_unet():
unet(x)
with pytest.raises(AssertionError):
# Check if num_stages matchs strides, len(enc_dilations)=num_stages
# Check if num_stages matches strides, len(enc_dilations)=num_stages
unet = UNet(
in_channels=3,
base_channels=4,
@ -500,7 +500,7 @@ def test_unet():
unet(x)
with pytest.raises(AssertionError):
# Check if num_stages matchs strides, len(dec_dilations)=num_stages-1
# Check if num_stages matches strides, len(dec_dilations)=num_stages-1
unet = UNet(
in_channels=3,
base_channels=4,

10
tests/test_models/test_utils/test_embed.py
View File

@ -173,7 +173,7 @@ def test_patch_embed():
# test L = out_h * out_w
assert shape[0] * shape[1] == x3.shape[1]
# test thte init_out_size with nn.Unfold
# test the init_out_size with nn.Unfold
assert patch_merge_3.init_out_size[1] == (input_size[0] - 2 * 4 -
1) // 2 + 1
assert patch_merge_3.init_out_size[0] == (input_size[0] - 2 * 4 -
@ -195,7 +195,7 @@ def test_patch_embed():
_, shape = patch_merge_3(dummy_input)
# when input_size equal to real input
# the out_size shoule be equal to `init_out_size`
# the out_size should be equal to `init_out_size`
assert shape == patch_merge_3.init_out_size
input_size = (H, W)
@ -213,7 +213,7 @@ def test_patch_embed():
_, shape = patch_merge_3(dummy_input)
# when input_size equal to real input
# the out_size shoule be equal to `init_out_size`
# the out_size should be equal to `init_out_size`
assert shape == patch_merge_3.init_out_size
# test adap padding
@ -288,7 +288,7 @@ def test_patch_embed():
assert out_size == (2, 1)
assert x_out.size(1) == out_size[0] * out_size[1]
# test different kernel_size with diffrent stride
# test different kernel_size with different stride
input_size = (6, 5)
kernel_size = (6, 2)
stride = (6, 2)
@ -437,7 +437,7 @@ def test_patch_merging():
assert out_size == (2, 1)
assert x_out.size(1) == out_size[0] * out_size[1]
# test different kernel_size with diffrent stride
# test different kernel_size with different stride
input_size = (6, 5)
kernel_size = (6, 2)
stride = (6, 2)

6
tools/model_converters/mit2mmseg.py
View File

@ -14,14 +14,14 @@ def convert_mit(ckpt):
for k, v in ckpt.items():
if k.startswith('head'):
continue
# patch embedding convertion
# patch embedding conversion
elif k.startswith('patch_embed'):
stage_i = int(k.split('.')[0].replace('patch_embed', ''))
new_k = k.replace(f'patch_embed{stage_i}', f'layers.{stage_i-1}.0')
new_v = v
if 'proj.' in new_k:
new_k = new_k.replace('proj.', 'projection.')
# transformer encoder layer convertion
# transformer encoder layer conversion
elif k.startswith('block'):
stage_i = int(k.split('.')[0].replace('block', ''))
new_k = k.replace(f'block{stage_i}', f'layers.{stage_i-1}.1')
@ -45,7 +45,7 @@ def convert_mit(ckpt):
new_k = new_k.replace('dwconv.dwconv.', '1.')
new_k = new_k.replace('fc2.', '4.')
string += f'{new_k} {v.shape}-{new_v.shape}'
# norm layer convertion
# norm layer conversion
elif k.startswith('norm'):
stage_i = int(k.split('.')[0].replace('norm', ''))
new_k = k.replace(f'norm{stage_i}', f'layers.{stage_i-1}.2')

4
tools/onnx2tensorrt.py
View File

@ -117,7 +117,7 @@ def onnx2tensorrt(onnx_file: str,
import tensorrt as trt
min_shape = input_config['min_shape']
max_shape = input_config['max_shape']
# create trt engine and wraper
# create trt engine and wrapper
opt_shape_dict = {'input': [min_shape, min_shape, max_shape]}
max_workspace_size = get_GiB(workspace_size)
trt_engine = onnx2trt(
@ -254,7 +254,7 @@ if __name__ == '__main__':
'Dataset {} does not found.'.format(args.dataset)
for max_value, min_value in zip(args.max_shape, args.min_shape):
assert max_value >= min_value, \
'max_shape sould be larger than min shape'
'max_shape should be larger than min shape'
input_config = {
'min_shape': args.min_shape,

2
tools/pytorch2torchscript.py
View File

@ -113,7 +113,7 @@ def pytorch2libtorch(model,
imgs = mm_inputs.pop('imgs')
# replace the orginal forword with forward_dummy
# replace the original forword with forward_dummy
model.forward = model.forward_dummy
model.eval()
traced_model = torch.jit.trace(