TransPose: Keypoint Localization via Transformer
              Sen Yang Zhibin Quan Mu Nie Wankou Yang*
           School of Automation, Southeast University, Nanjing 210096, China
               {yangsenius, 101101872, niemu, wkyang}@seu.edu.cn

            Abstract
  While CNN-based models have made remarkable
progress on human pose estimation, what spatial depen-
dencies they capture to localize keypoints remains un-
clear. In this work, we propose a model called Trans-
Pose, which introduces ...