Awesome

Context-Aware-CF-Tracking

Authors: Matthias Mueller, Neil Smith and Bernard Ghanem

License: See LICENSE file

If you use any of this work please cite:
Matthias Mueller, Neil Smith, Bernard Ghanem
"Context-Aware Correlation Filter Tracking"
Conference on Computer Vision and Pattern Recognition (CVPR 2017) [Oral]

Running trackers on test sequences

Every trackers comes with one sequence from OTB100. Simply run the function run_tracker. If you would like to run a different sequence copy its folder into the sequences folder. Then change the video name in the run_tracker function.

Running trackers on OTB benchmark

Copy the trackers into the folder trackers of the benchmark. Add the following to the config_trackers function.

 trackersCA = {struct('name','MOSSE_CA','namePaper','MOSSE_CA'),...     
     struct('name','DCF_CA','namePaper','DCF_CA'),...   
     struct('name','SAMF_CA','namePaper','SAMF_CA'),...
     struct('name','STAPLE_CA','namePaper','STAPLE_CA')};

trackers = trackersCA;

Looking for a large-scale dataset for tracking?

Check out TrackingNet. Instructions on how to download the dataset can be found here.