Attention-Based 3D-CNNs for Large-Vocabulary Sign Language Recognition | IEEE Journals & Magazine | IEEE Xplore