“…To evaluate the superiority of our proposed method, we compare our method with some existing state-of-the-art RGB-T trackers, including MANet [17], DAFNet [6], Net [39], TODA [28], MACNet [30], CAT [12], CEDiMP [38], SiamCDA [36], mfDiMP [31], FANet [40], CBPNet [27], MANet++ [33], JMMAC [32], ADRNet [33], TFNet [41], MFGNet [25], M5LNet [23], HMFT [34], APFNet [26] and MFNet [35], on three challenging datasets. Considering that existing methods usually employ different training datasets, we use the LasHeR training set to retrain some of these algorithms for fair comparisons, including FANet * , DAFNet * , MANet * and mfDiMP * .…”