Fully-Convolutional Siamese Networks for Object Tracking

30 Jun 2016 • Luca Bertinetto • Jack Valmadre • João F. Henriques • Andrea Vedaldi • Philip H. S. Torr

The problem of arbitrary object tracking has traditionally been tackled by learning a model of the object's appearance exclusively online, using as sole training data the video itself. Despite the success of these methods, their online-only approach inherently limits the richness of the model they can learn... (read more)

PDF Abstract

Results from the Paper


 SOTA for Visual Object Tracking on DAVIS 2016 (Jaccard (Decay) metric )

     Get a GitHub badge
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT LEADERBOARD
Visual Object Tracking DAVIS 2016 + Jaccard (Recall) 86.0 # 14
Jaccard (Decay) 0.8 # 1
F-measure (Mean) 67.1 # 14
F-measure (Recall) 79.0 # 13
F-measure (Decay) 0.3 # 1

Results from Other Papers


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK SOURCE PAPER COMPARE
Visual Object Tracking OTB-2013 SiamFC-3s AUC 0.607 # 6
Visual Object Tracking OTB-50 SiamFC-3s AUC 0.516 # 3