Superimposed-Masked-Dataset.py

# This script was modified from the imagenet-1k HF dataset repo: https://huggingface.co/datasets/imagenet-1k

import os
import numpy as np

import datasets
from datasets.tasks import ImageClassification

from .classes import IMAGENET2012_CLASSES
from io import BytesIO


_CITATION = """\
@article{BibTeX
}
"""

_HOMEPAGE = "https://arielnlee.github.io/PatchMixing/"

_DESCRIPTION = """\
SMD is an occluded ImageNet-1K validation set, created to be an additional way to evaluate the impact of occlusion on model performance. This experiment used a variety of occluder objects that are not in the ImageNet-1K label space and are unambiguous in relationship to objects that reside in the label space.
"""

_DATA_URL = {
    "smd": [
        f"https://huggingface.co/datasets/ariellee/Superimposed-Masked-Dataset/resolve/main/smd_{i}.tar.gz"
        for i in range(1, 41)
    ]
}

_MASK_DATA_URL = {
    "smd_masks": [
        f"https://huggingface.co/datasets/ariellee/Superimposed-Masked-Dataset/resolve/main/SMD_masks.tar.gz"
    ]
}

class SMD(datasets.GeneratorBasedBuilder):
    VERSION = datasets.Version("1.0.0")

    DEFAULT_WRITER_BATCH_SIZE = 1000

    def _info(self):
        assert len(IMAGENET2012_CLASSES) == 1000
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "image": datasets.Image(),
                    "label": datasets.ClassLabel(names=list(IMAGENET2012_CLASSES.values())),
                    "segmentation": datasets.Sequence(datasets.Array2D(shape=(None, None), dtype="float32"))
                }
            ),
            homepage=_HOMEPAGE,
            citation=_CITATION,
            task_templates=[ImageClassification(image_column="image", label_column="label")],
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        archives = dl_manager.download_and_extract(_DATA_URL)
        mask_archives = dl_manager.download_and_extract(_MASK_DATA_URL)
    
        return [
            datasets.SplitGenerator(
                name="SMD",
                gen_kwargs={
                    "archives": archives["smd"],
                    "mask_archives": mask_archives["smd_masks"],
                },
            ),
        ]
    
    def _generate_examples(self, archives, mask_archives):
        """Yields examples."""
        idx = 0
        mask_files = {}
        for mask_archive in mask_archives:
            for path, file in dl_manager.iter_archive(mask_archive):
                if path.endswith(".npy"):
                    mask_files[path] = np.load(BytesIO(file.read()))
        
        for archive in archives:
            for path, file in dl_manager.iter_archive(archive):
                if path.endswith(".png"):
                    synset_id = os.path.basename(os.path.dirname(path))
                    label = IMAGENET2012_CLASSES[synset_id]
        
                    mask_file_path = path.replace("_occluded.png", "_mask.npy")
                    segmentation_mask = mask_files.get(mask_file_path, None)
                    if segmentation_mask is not None:
                        ex = {
                            "image": {"path": path, "bytes": file.read()}, 
                            "label": label, 
                            "segmentation": segmentation_mask.tolist()  # Convert numpy array to list
                        }
                        yield idx, ex
                        idx += 1