1 dataset found

Groups: Cross-Modal Learning Formats: JSON

Filter Results

  • Places

    The dataset used in the paper is Places, a large dataset of 400k pairs of images from the Places 205 dataset and corresponding spoken audio captions.