Data Process¶
In MMClassification, the data process and the dataset is decomposed. The datasets only define how to get samples’ basic information from the file system. These basic information includes the ground-truth label and raw images data / the paths of images.The data process includes data transforms, data preprocessors and batch augmentations.
Data Transforms
: Transforms includes loading, preprocessing, formatting and etc.Data Preprocessors
: Processes includes collate, normalization, stacking, channel fliping and etc.Batch Augmentations
: Batch augmentation involves multiple samples, such as Mixup and CutMix.
Data Transforms¶
To prepare the inputs data, we need to do some transforms on these basic
information. These transforms includes loading, preprocessing and
formatting. And a series of data transforms makes up a data pipeline.
Therefore, you can find the a pipeline
argument in the configs of dataset,
for example:
train_pipeline = [
dict(type='LoadImageFromFile'),
dict(type='RandomResizedCrop', scale=224),
dict(type='RandomFlip', prob=0.5, direction='horizontal'),
dict(type='PackClsInputs'),
]
train_dataloader = dict(
....
dataset=dict(
pipeline=train_pipeline,
....),
....
)
Every item of a pipeline list is one of the following data transforms class. And if you want to add a custom data transformation class, the tutorial Custom Data Pipelines will help you.
Processing and Augmentation¶
Wrapper to use augmentation from albumentations library. |
|
Randomly change the brightness, contrast and saturation of an image. |
|
EfficientNet style center crop. |
|
EfficientNet style RandomResizedCrop. |
|
Adjust images lighting using AlexNet-style PCA jitter. |
|
Crop the given Image at a random location. |
|
Randomly selects a rectangle region in an image and erase pixels. |
|
Crop the given image to random scale and aspect ratio. |
|
Resize images along the specified edge. |
Composed Augmentation¶
Composed augmentation is a kind of methods which compose a series of data
augmentation transforms, such as AutoAugment
and RandAugment
.
Auto augmentation. |
|
Random augmentation. |
To specify the augmentation combination (The policies
argument), you can use string to specify
from some preset policies.
Preset policy |
Use for |
Description |
---|---|---|
“imagenet” |
Policy for ImageNet, come from DeepVoltaire/AutoAugment |
|
“timm_increasing” |
The |
And you can also configure a group of policies manually by selecting from the below table.
Auto adjust image contrast. |
|
Adjust images brightness. |
|
Adjust images color balance. |
|
Adjust images contrast. |
|
Cutout images. |
|
Equalize the image histogram. |
|
Invert images. |
|
Posterize images (reduce the number of bits for each color channel). |
|
Rotate images. |
|
Adjust images sharpness. |
|
Shear images. |
|
Solarize images (invert all pixel values above a threshold). |
|
SolarizeAdd images (add a certain value to pixels below a threshold). |
|
Translate images. |
|
The base class of augmentation transform for RandAugment. |
Formatting¶
Collect and only reserve the specified fields. |
|
Pack the inputs data for the classification. |
|
Convert object to |
|
Convert the image from OpenCV format to |
|
Transpose numpy array. |
MMCV transforms¶
We also provides many transforms in MMCV. You can use them directly in the config files. Here are some frequently used transforms, and the whole transforms list can be found in mmcv.transforms.
Load an image from file. |
|
Resize images & bbox & seg & keypoints. |
|
Random resize images & bbox & keypoints. |
|
Flip the image & bbox & keypoints & segmentation map. |
|
Randomly convert image to grayscale with a probability. |
|
Crop the center of the image, segmentation masks, bounding boxes and key points. If the crop area exceeds the original image and |
|
Normalize the image. |
|
Compose multiple transforms sequentially. |
Data Preprocessors¶
The data preprocessor is also a component to process the data before feeding data to the neural network. Comparing with the data transforms, the data preprocessor is a module of the classifier, and it takes a batch of data to process, which means it can use GPU and batch to accelebrate the processing.
The default data preprocessor in MMClassification could do the pre-processing like following:
Move data to the target device.
Pad inputs to the maximum size of current batch.
Stack inputs to a batch.
Convert inputs from bgr to rgb if the shape of input is (3, H, W).
Normalize image with defined std and mean.
Do batch augmentations like Mixup and CutMix during training.
You can configure the data preprocessor by the data_preprocessor
field or model.data_preprocessor
field in the config file. Typical usages are as below:
data_preprocessor = dict(
# RGB format normalization parameters
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True, # convert image from BGR to RGB
)
Or define in model.data_preprocessor
as following:
model = dict(
backbone = ...,
neck = ...,
head = ...,
data_preprocessor = dict(
mean=[123.675, 116.28, 103.53],
std=[58.395, 57.12, 57.375],
to_rgb=True)
train_cfg=...,
)
Note that the model.data_preprocessor
has higher priority than data_preprocessor
.
Image pre-processor for classification tasks. |
Batch Augmentations¶
The batch augmentation is a component of data preprocessors. It involves multiple samples and mix them in some way, such as Mixup and CutMix.
These augmentations are usually only used during training, therefore, we use the model.train_cfg
field to configure them in config files.
model = dict(
backbone=...,
neck=...,
head=...,
train_cfg=dict(augments=[
dict(type='Mixup', alpha=0.8),
dict(type='CutMix', alpha=1.0),
]),
)
You can also specify the probabilities of every batch augmentation by the probs
field.
model = dict(
backbone=...,
neck=...,
head=...,
train_cfg=dict(augments=[
dict(type='Mixup', alpha=0.8),
dict(type='CutMix', alpha=1.0),
], probs=[0.3, 0.7])
)
Here is a list of batch augmentations can be used in MMClassification.
Mixup batch augmentation. |
|
CutMix batch agumentation. |
|
ResizeMix Random Paste layer for a batch of data. |