1 dataset found

Groups: Speech Recognition Organizations: No Organization

Filter Results

  • BABEL dataset

    The dataset used in this paper is the BABEL dataset, which contains 10881 motion sequences, with 65926 subsequences and the corresponding textual labels.