ex-dark.py

# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Exclusively Dark Image Dataset"""


import os

import datasets
import pandas as pd


_CITATION = """\
@article{Exdark,
  title = {Getting to Know Low-light Images with The Exclusively Dark Dataset},
  author = {Loh, Yuen Peng and Chan, Chee Seng},
  journal = {Computer Vision and Image Understanding},
  volume = {178},
  pages = {30-42},
  year = {2019},
  doi = {https://doi.org/10.1016/j.cviu.2018.10.010}
}
"""

_DESCRIPTION = """\
The Exclusively Dark (ExDARK) dataset is a collection of low-light
images from very low-light environments to twilight (i.e 10 different
conditions) with 12 object classes (similar to PASCAL VOC) annotated on both
image class level and local object bounding boxes.

The object classes are as follows:

- Dog
- Motorbike
- People
- Cat
- Chair
- Table
- Car
- Bicycle
- Bottle
- Bus
- Cup
- Boat

For more information about the original Exclusively Dark Image dataset,
please visit the official dataset page:

https://github.com/cs-chan/Exclusively-Dark-Image-Dataset

Please refer to the original dataset source for any additional details,
citations, or specific usage guidelines provided by the dataset creators.
"""

_HOMEPAGE = "https://github.com/cs-chan/Exclusively-Dark-Image-Dataset"

_LICENSE = "bsd-3-clause"

_LABEL_NAMES = [
    "Dog",
    "Motorbike",
    "People",
    "Cat",
    "Chair",
    "Table",
    "Car",
    "Bicycle",
    "Bottle",
    "Bus",
    "Cup",
    "Boat",
]


class ExDark(datasets.GeneratorBasedBuilder):
    """Exclusively Dark (ExDARK) dataset"""

    VERSION = datasets.Version("1.0.0")

    BUILDER_CONFIGS = [
        datasets.BuilderConfig(
            name="exdark",
            version=VERSION,
            description="Exclusively Dark (ExDARK) dataset",
        ),
    ]

    DEFAULT_CONFIG_NAME = "exdark"

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "img": datasets.Image(),
                    "target": datasets.Sequence(
                        feature={
                            "labels": datasets.features.ClassLabel(
                                names=_LABEL_NAMES
                            ),
                            "bboxes": datasets.features.Sequence(
                                feature=datasets.Value("float32")
                            )
                        }
                    )
                }
            ),
            homepage=_HOMEPAGE,
            license=_LICENSE,
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        data_dir = dl_manager.download_and_extract("ExDark.zip")

        metadata_path = os.path.join(data_dir, "ExDark", "metadata.csv")

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "data_dir": data_dir,
                    "metadata_path": metadata_path,
                    "split": "train",
                },
            ),
        ]

    def _generate_examples(self, data_dir, metadata_path, split):
        df = pd.read_csv(metadata_path)
        classes = df["class"].unique()

        df["class"] = df["class"].apply(lambda x: classes.tolist().index(x))

        for idx, file_name in enumerate(df.file_name.unique()):
            img_path = os.path.join(data_dir, "ExDark", file_name)
            sample = df[df.file_name == file_name]
            bboxs = sample[["x", "y", "w", "h"]].to_numpy()
            labels = sample["class"].to_numpy()
            yield idx, {
                "img": img_path,
                "target": {
                    "labels": labels,
                    "bboxes": bboxs,
                },
            }