charades.py

# coding=utf-8
# Copyright 2022 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.

# Lint as: python3
"""Charades is dataset composed of 9848 videos of daily indoors activities collected through Amazon Mechanical Turk"""


import csv
import os

import datasets

from .classes import CHARADES_CLASSES

_CITATION = """
@article{sigurdsson2016hollywood,
    author = {Gunnar A. Sigurdsson and G{\"u}l Varol and Xiaolong Wang and Ivan Laptev and Ali Farhadi and Abhinav Gupta},
    title = {Hollywood in Homes: Crowdsourcing Data Collection for Activity Understanding},
    journal = {ArXiv e-prints},
    eprint = {1604.01753},
    year = {2016},
    url = {http://arxiv.org/abs/1604.01753},
}
"""

_DESCRIPTION = """\
Charades is dataset composed of 9848 videos of daily indoors activities collected through Amazon Mechanical Turk. 267 different users were presented with a sentence, that includes objects and actions from a fixed vocabulary, and they recorded a video acting out the sentence (like in a game of Charades). The dataset contains 66,500 temporal annotations for 157 action classes, 41,104 labels for 46 object classes, and 27,847 textual descriptions of the videos.
"""


_ANNOTATIONS_URL = "https://ai2-public-datasets.s3-us-west-2.amazonaws.com/charades/Charades.zip"
_VIDEOS_URL = {
    "default": "https://ai2-public-datasets.s3-us-west-2.amazonaws.com/charades/Charades_v1.zip",
    "480p": "https://ai2-public-datasets.s3-us-west-2.amazonaws.com/charades/Charades_v1_480.zip",
}


class Charades(datasets.GeneratorBasedBuilder):
    """Charades is dataset composed of 9848 videos of daily indoors activities collected through Amazon Mechanical Turk"""

    BUILDER_CONFIGS = [datasets.BuilderConfig(name="default"), datasets.BuilderConfig(name="480p")]
    DEFAULT_CONFIG_NAME = "default"

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "video_id": datasets.Value("string"),
                    "video": datasets.Value("string"),
                    "subject": datasets.Value("string"),
                    "scene": datasets.Value("string"),
                    "quality": datasets.Value("int32"),
                    "relevance": datasets.Value("int32"),
                    "verified": datasets.Value("string"),
                    "script": datasets.Value("string"),
                    "objects": datasets.features.Sequence(datasets.Value("string")),
                    "descriptions": datasets.features.Sequence(datasets.Value("string")),
                    "labels": datasets.Sequence(
                        datasets.features.ClassLabel(
                            num_classes=len(CHARADES_CLASSES), names=list(CHARADES_CLASSES.values())
                        )
                    ),
                    "action_timings": datasets.Sequence(datasets.Sequence(datasets.Value("float32"))),
                    "length": datasets.Value("float32"),
                }
            ),
            supervised_keys=None,
            homepage="",
            citation=_CITATION,
        )

    def _split_generators(self, dl_manager):
        annotations_path = dl_manager.download_and_extract(_ANNOTATIONS_URL)
        archive = os.path.join(dl_manager.download_and_extract(_VIDEOS_URL[self.config.name]), "Charades_v1")
        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN,
                gen_kwargs={
                    "annotation_file": os.path.join(annotations_path, "Charades", "Charades_v1_train.csv"),
                    "video_folder": archive,
                },
            ),
            datasets.SplitGenerator(
                name=datasets.Split.TEST,
                gen_kwargs={
                    "annotation_file": os.path.join(annotations_path, "Charades", "Charades_v1_test.csv"),
                    "video_folder": archive,
                },
            ),
        ]

    def _generate_examples(self, annotation_file, video_folder):
        """This function returns the examples."""
        with open(annotation_file, "r", encoding="utf-8") as csv_file:
            reader = csv.DictReader(csv_file)
            idx = 0
            for row in reader:
                path = os.path.join(video_folder, row["id"] + ".mp4")
                labels = []
                action_timings = []
                for class_label in row["actions"].split(";"):
                    # Skip empty action labels
                    if len(class_label) != 0:
                        # format is like: "c123 11.0 13.0"
                        labels.append(CHARADES_CLASSES[class_label.split(" ")[0]])
                        timings = list(map(float, class_label.split(" ")[1:]))
                        action_timings.append(timings)

                yield idx, {
                    "video_id": row["id"],
                    "video": path,
                    "subject": row["subject"],
                    "scene": row["scene"],
                    "quality": int(row["quality"]) if len(row["quality"]) != 0 else -100,
                    "relevance": int(row["relevance"]) if len(row["relevance"]) != 0 else -100,
                    "verified": row["verified"],
                    "script": row["script"],
                    "objects": row["objects"].split(";"),
                    "descriptions": row["descriptions"].split(";"),
                    "labels": labels,
                    "action_timings": action_timings,
                    "length": row["length"],
                }

                idx += 1