vqa_benchmarking_backend.datasets.CLEVRDataset

Module Contents

Classes

CLEVRDataSample

Class describing one data sample of the CLEVR dataset

CLEVRDataset

Class describing the CLEVR dataset

Functions

preprocess_question(question: str) → List[str]

Remove punctuation and make everything lower case

load_img(path: str, transform=None) → numpy.ndarray

Load an image using module cv2

load_img_feats(path: str) → torch.FloatTensor

Load a numpy array containing image features
vqa_benchmarking_backend.datasets.CLEVRDataset.preprocess_question(question: str) → List[str]

Remove punctuation and make everything lower case

vqa_benchmarking_backend.datasets.CLEVRDataset.load_img(path: str, transform=None) → numpy.ndarray

Load an image using module cv2

vqa_benchmarking_backend.datasets.CLEVRDataset.load_img_feats(path: str) → torch.FloatTensor

Load a numpy array containing image features

class vqa_benchmarking_backend.datasets.CLEVRDataset.CLEVRDataSample(question_id: str, question: str, answers: Dict[str, float], image_id: str, image_path: str, image_feat_path: str, image_transform=None)

Bases: vqa_benchmarking_backend.datasets.dataset.DataSample

Class describing one data sample of the CLEVR dataset Inheriting from DataSample

property image(self) → numpy.ndarray

Returns the image, if not present it loads it from self._image_path

property question_tokenized(self) → List[str]

Returns tokenized question

property question(self) → str

Returns full question

__str__(self)

Stringify object

question_token_ids(self, vocab: vqa_benchmarking_backend.utils.vocab.Vocabulary) → torch.LongTensor
class vqa_benchmarking_backend.datasets.CLEVRDataset.CLEVRDataset(question_file: str, img_dir, img_feat_dir, idx2ans, transform=None, load_img_features=False, dataset_fraction: float = 0.05, random_seed: int = 12345)

Bases: vqa_benchmarking_backend.datasets.dataset.DiagnosticDataset

Class describing the CLEVR dataset Inheriting from DiagnosticDataset

_load_data(self, question_file: str, dataset_fraction, random_seed) → Tuple[List[vqa_benchmarking_backend.datasets.dataset.DataSample], Dict[str, vqa_benchmarking_backend.datasets.dataset.DataSample], vqa_benchmarking_backend.utils.vocab.Vocabulary, vqa_benchmarking_backend.utils.vocab.Vocabulary]

Loads data from CLEVR json files Returns:

  • data: list of CLEVRDataSample

  • qid_to_sample: mapping of question id to data sample

  • question_vocab: Vocabulary of all unique words occuring in the data

  • answer_vocab: Vocabulary of all unique answers

__getitem__(self, index) → vqa_benchmarking_backend.datasets.dataset.DataSample

Returns a data sample

label_from_class(self, class_index: int) → str

Get the answer string of a given class index

word_in_vocab(self, word: str) → bool

Checks if a word occured inside the Vocabulary dervied of all questions

get_name(self) → str

Returns the name of the dataset, required for file caching

__len__(self)

Returns the length of the CLEVRDataset as in self.data

index_to_question_id(self, index) → str

Get the index of a specific question id

class_idx_to_answer(self, class_idx: int) → Union[str, None]

Get the answer string for a given class index from the self.idx2ans dictionary