TY - RPRT TI - Self-Supervised Joint Encoding of Motion and Appearance for First Person Action Recognition AU - Mirco Planamente AU - Andrea Bottino AU - Barbara Caputo PY - 2020 UR - https://arxiv.org/abs/2002.03982 ID - 2002.03982 ER -