search

foolwood / DCFNet_pytorch

DCFNet: Discriminant Correlation Filters Network for Visual Tracking
https://arxiv.org/pdf/1704.04057.pdf
MIT License
208 stars 60 forks source link

Is the AUC scored 0.665 obtained using this set of parameters? #11

Open LCAR979 opened 6 years ago

LCAR979 commented 6 years ago

By keeping your default parameters, without fine-tune, I got 0.6466 on OTB2013 and 0.6206 on OTB2015 (guess you are using py2 so I also tested using py2). So is that 0.665 score obtained by matlab version?

shallowtoil commented 6 years ago

By keeping your default parameters, without fine-tune, I got 0.6466 on OTB2013 and 0.6206 on OTB2015 (guess you are using py2 so I also tested using py2). So is that 0.665 score obtained by matlab version?

why I just got 0.6248 success rate on OTB2013 and 0.6037 on OTB2015？ Have you changed the python version? How's it going?

LCAR979 commented 6 years ago

@jensenzhoujh, I got that result using py2. And I also tested if changed to py3, a few compatible problems will occur

shallowtoil commented 5 years ago

@LCAR979 I changed to python 2.7 but got 0.6336/0.6014. Could you offer more details about the package versions?

ucasqcz commented 5 years ago

By keeping your default parameters, without fine-tune, I got 0.6466 on OTB2013 and 0.6206 on OTB2015 (guess you are using py2 so I also tested using py2). So is that 0.665 score obtained by matlab version?

why I just got 0.6248 success rate on OTB2013 and 0.6037 on OTB2015？ Have you changed the python version? How's it going?

why do the results change when the python version changes?

LCAR979 commented 5 years ago

@jensenzhoujh I tested using python2.7.15 with pytorch0.4.1, under commit #17007f . @ucasqcz Division behaves differently in py2 and py3 and I think the author codes towards py2 ( there exists some compatibility problems)

SY-Xuan commented 5 years ago

I also got the 0.6466 success rate on OTB2013 too with python2.7