TY - RPRT TI - DropMAE: Learning Representations via Masked Autoencoders with Spatial-Attention Dropout for Temporal Matching Tasks AU - Qiangqiang Wu AU - Tianyu Yang AU - Ziquan Liu AU - Wei Lin AU - Baoyuan Wu AU - Antoni B. Chan PY - 2025 UR - https://arxiv.org/abs/2304.00571 ID - 2304.00571 ER -