1 paper
Yueyang Cang, Xiaoteng Zhang, Zhiyuan Ning +2
Visual object tracking requires effective temporal integration, yet most Transformer trackers still rely on predominantly feed-forward feature extraction. Existing temporal mechani…