1 dataset found

Groups: Human Action Recognition Formats: JSON

Filter Results

  • AVA

    The dataset used in this paper is a Flickr image dataset, which is used to evaluate the proposed deep aesthetic feature learning framework.