super-resolution-games.py

import datasets
import cv2


_DESCRIPTION = ''

_PROJECTS_GAME_ENGINE = [
    'GameEngine_ActionRPG',
    'GameEngine_ArchVizInterior',
    'GameEngine_ASCTeuthisan',
    'GameEngine_BroadcastSample',
    'GameEngine_CitySample',
    'GameEngine_ElectricDreamsEnv',
    'GameEngine_ElementalDemo',
    'GameEngine_HillsideSample',
    'GameEngine_Matinee',
    'GameEngine_MeerkatDemo',
    'GameEngine_MLDeformerSample',
    'GameEngine_ParticleEffects',
    'GameEngine_RealisticRendering',
    'GameEngine_SlayAnimationSample',
    'GameEngine_StylizedRendering',
    'GameEngine_SubwaySequencer',
    'GameEngine_SunTemple'
]

_PROJECTS_DOWNSCALE = [
    'Downscale_Dota2'
]

# _PROJECTS = _PROJECTS_GAME_ENGINE + _PROJECTS_DOWNSCALE
_PROJECTS = _PROJECTS_DOWNSCALE

_DESCRIPTION_DATA_GAME_ENGINE = {
    'GameEngine_ActionRPG': '',
    'GameEngine_ArchVizInterior': '',
    'GameEngine_ASCTeuthisan': '',
    'GameEngine_BroadcastSample': '',
    'GameEngine_CitySample': '',
    'GameEngine_ElectricDreamsEnv': '',
    'GameEngine_ElementalDemo': '',
    'GameEngine_HillsideSample': '',
    'GameEngine_Matinee': '',
    'GameEngine_MeerkatDemo': '',
    'GameEngine_MLDeformerSample': '',
    'GameEngine_ParticleEffects': '',
    'GameEngine_RealisticRendering': '',
    'GameEngine_SlayAnimationSample': '',
    'GameEngine_StylizedRendering': '',
    'GameEngine_SubwaySequencer': '',
    'GameEngine_SunTemple': ''
}

_DESCRIPTION_DATA_DOWNSCALE = {
    'Downscale_Dota2': ''
}

_DESCRIPTION_DATA = _DESCRIPTION_DATA_DOWNSCALE

# _DESCRIPTION_DATA = {
#     **_DESCRIPTION_DATA_GAME_ENGINE,
#     **_DESCRIPTION_DATA_DOWNSCALE
# }

_DATA_FILES_GAME_ENGINE = {
    'GameEngine_ActionRPG': '',
    'GameEngine_ArchVizInterior': '',
    'GameEngine_ASCTeuthisan': '',
    'GameEngine_BroadcastSample': '',
    'GameEngine_CitySample': '',
    'GameEngine_ElectricDreamsEnv': '',
    'GameEngine_ElementalDemo': '',
    'GameEngine_HillsideSample': '',
    'GameEngine_Matinee': '',
    'GameEngine_MeerkatDemo': '',
    'GameEngine_MLDeformerSample': '',
    'GameEngine_ParticleEffects': '',
    'GameEngine_RealisticRendering': '',
    'GameEngine_SlayAnimationSample': '',
    'GameEngine_StylizedRendering': '',
    'GameEngine_SubwaySequencer': '',
    'GameEngine_SunTemple': ''
}

_DATA_FILES_DOWNSCALE = {
    'Downscale_Dota2': {
        'train': {
            'r270p': 'data/DownscaleData/Dota2/train-270p.tar.gz',
            'r360p': 'data/DownscaleData/Dota2/train-360p.tar.gz',
            'r540p': 'data/DownscaleData/Dota2/train-540p.tar.gz',
            'r1080p': 'data/DownscaleData/Dota2/train-1080p.tar.gz'
        },
        'val': {
            'r270p': 'data/DownscaleData/Dota2/val-270p.tar.gz',
            'r360p': 'data/DownscaleData/Dota2/val-360p.tar.gz',
            'r540p': 'data/DownscaleData/Dota2/val-540p.tar.gz',
            'r1080p': 'data/DownscaleData/Dota2/val-1080p.tar.gz'
        }
    }
}

# _DATA_FILES = {
#     **_DATA_FILES_GAME_ENGINE,
#     **_DATA_FILES_DOWNSCALE
# }

_DATA_FILES = _DATA_FILES_DOWNSCALE


class SuperResolutionGamesConfig(datasets.BuilderConfig):
    def __init__(self, data_files, features, **kwargs):
        super().__init__(version=datasets.Version('0.0.0'), **kwargs)
        self.data_files = data_files
        self.features = features


class SuperResolutionGames(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        SuperResolutionGamesConfig(
            name=name,
            description=_DESCRIPTION_DATA[name],
            data_files=_DATA_FILES[name],
            features=['r270p', 'r360p', 'r540p', 'r1080p']
        ) for name in _PROJECTS
    ]

    def _info(self):
        features = {f: datasets.Image() for f in self.config.features}
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(features),
            supervised_keys=None,
        )

    def _split_generators(self, dl_manager):
        data_files = self.config.data_files
        train_archives, val_archives = data_files['train'], data_files['val']

        train_archives_downloaded = {
            k: dl_manager.download(train_archives[k]) \
                for k in train_archives.keys()
        }

        val_archives_downloaded = {
            k: dl_manager.download(val_archives[k]) \
                for k in val_archives.keys()
        }

        splits = [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    k: dl_manager.iter_archive(train_archives_downloaded[k]) \
                        for k in train_archives_downloaded.keys()
                }
            ),
            datasets.SplitGenerator(
                name=datasets.Split.VALIDATION,
                gen_kwargs={
                    k: dl_manager.iter_archive(val_archives_downloaded[k]) \
                        for k in val_archives_downloaded.keys()
                }
            )
        ]
        return splits

    def _generate_examples(self, r270p, r360p, r540p, r1080p):
        for idx, resolutions in enumerate(zip(r270p, r360p, r540p, r1080p)):
            r270p_i, r360p_i, r540p_i, r1080p_i = resolutions
            sample = {
                'r270p': {'path': r270p_i[0], 'bytes': r270p_i[1].read()},
                'r360p': {'path': r360p_i[0], 'bytes': r360p_i[1].read()},
                'r540p': {'path': r540p_i[0], 'bytes': r540p_i[1].read()},
                'r1080p': {'path': r1080p_i[0], 'bytes': r1080p_i[1].read()}
            }
            yield idx, sample