RPAN: An End-to-End Recurrent Pose-Attention Network for Action Recognition in Videos
Distribution of the number of citations over years.