Source code for torchgeo.datamodules.resisc45
# Copyright (c) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License.
"""RESISC45 datamodule."""
from typing import Any, Dict, Optional
import kornia.augmentation as K
import matplotlib.pyplot as plt
import pytorch_lightning as pl
import torch
from torch.utils.data import DataLoader
from torchvision.transforms import Compose, Normalize
from ..datasets import RESISC45
# https://github.com/pytorch/pytorch/issues/60979
# https://github.com/pytorch/pytorch/pull/61045
DataLoader.__module__ = "torch.utils.data"
class RESISC45DataModule(pl.LightningDataModule):
"""LightningDataModule implementation for the RESISC45 dataset.
Uses the train/val/test splits from the dataset.
"""
band_means = torch.tensor( # type: ignore[attr-defined]
[0.36801773, 0.38097873, 0.343583]
)
band_stds = torch.tensor( # type: ignore[attr-defined]
[0.14540215, 0.13558227, 0.13203649]
)
[docs] def __init__(
self, root_dir: str, batch_size: int = 64, num_workers: int = 0, **kwargs: Any
) -> None:
"""Initialize a LightningDataModule for RESISC45 based DataLoaders.
Args:
root_dir: The ``root`` arugment to pass to the RESISC45 Dataset classes
batch_size: The batch size to use in all created DataLoaders
num_workers: The number of workers to use in all created DataLoaders
"""
super().__init__() # type: ignore[no-untyped-call]
self.root_dir = root_dir
self.batch_size = batch_size
self.num_workers = num_workers
self.norm = Normalize(self.band_means, self.band_stds)
[docs] def on_after_batch_transfer(
self, batch: Dict[str, Any], batch_idx: int
) -> Dict[str, Any]:
"""Apply batch augmentations after batch is transferred to the device.
Args:
batch: mini-batch of data
batch_idx: batch index
Returns:
augmented mini-batch
"""
if (
hasattr(self, "trainer")
and hasattr(self.trainer, "training")
and self.trainer.training # type: ignore[union-attr]
):
x = batch["image"]
train_augmentations = K.AugmentationSequential(
K.RandomRotation(p=0.5, degrees=90),
K.RandomHorizontalFlip(p=0.5),
K.RandomVerticalFlip(p=0.5),
K.RandomSharpness(p=0.5),
K.RandomErasing(p=0.1),
K.ColorJitter(
p=0.5, brightness=0.1, contrast=0.1, saturation=0.1, hue=0.1
),
data_keys=["input"],
)
x = train_augmentations(x)
batch["image"] = x
return batch
[docs] def preprocess(self, sample: Dict[str, Any]) -> Dict[str, Any]:
"""Transform a single sample from the Dataset.
Args:
sample: input image dictionary
Returns:
preprocessed sample
"""
sample["image"] = sample["image"].float()
sample["image"] /= 255.0
sample["image"] = self.norm(sample["image"])
return sample
[docs] def prepare_data(self) -> None:
"""Make sure that the dataset is downloaded.
This method is only called once per run.
"""
RESISC45(self.root_dir, checksum=False)
[docs] def setup(self, stage: Optional[str] = None) -> None:
"""Initialize the main ``Dataset`` objects.
This method is called once per GPU per run.
Args:
stage: stage to set up
"""
transforms = Compose([self.preprocess])
self.train_dataset = RESISC45(self.root_dir, "train", transforms=transforms)
self.val_dataset = RESISC45(self.root_dir, "val", transforms=transforms)
self.test_dataset = RESISC45(self.root_dir, "test", transforms=transforms)
[docs] def train_dataloader(self) -> DataLoader[Any]:
"""Return a DataLoader for training.
Returns:
training data loader
"""
return DataLoader(
self.train_dataset,
batch_size=self.batch_size,
num_workers=self.num_workers,
shuffle=True,
)
[docs] def val_dataloader(self) -> DataLoader[Any]:
"""Return a DataLoader for validation.
Returns:
validation data loader
"""
return DataLoader(
self.val_dataset,
batch_size=self.batch_size,
num_workers=self.num_workers,
shuffle=False,
)
[docs] def test_dataloader(self) -> DataLoader[Any]:
"""Return a DataLoader for testing.
Returns:
testing data loader
"""
return DataLoader(
self.test_dataset,
batch_size=self.batch_size,
num_workers=self.num_workers,
shuffle=False,
)
[docs] def plot(self, *args: Any, **kwargs: Any) -> plt.Figure:
"""Run :meth:`torchgeo.datasets.RESISC45.plot`."""
return self.val_dataset.plot(*args, **kwargs)