Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Results on OTB-100 #20

Open
chenfsjz opened this issue Jun 25, 2021 · 3 comments
Open

Results on OTB-100 #20

chenfsjz opened this issue Jun 25, 2021 · 3 comments

Comments

@chenfsjz
Copy link

HI, I have test the performance of OTB-100 with the results you provide. While the maximum success of TrDiMP is 0.700 on TrDiMP_002, which is less than your paper.

@594422814
Copy link
Owner

Sorry for the late response. Which toolkit did you test the OTB benchmark? I tested the OTB performance using my own toolkit, but I also tested other trackers (e.g., DiMP, PrDiMP) and could match their results. I will check it later.

@chenfsjz
Copy link
Author

I test the results on official toolkits of OTB-100 (http://cvlab.hanyang.ac.kr/tracker_benchmark/datasets.html), the maximal AUC of OTB-100 among TrDiMP_001, TrDiMP_002, TrDiMP_003 is 0.700. In addition, I test the results using GOT-10k toolkit, the highest AUC is 0.701 for TrDiMP_002.

@594422814
Copy link
Owner

I check again, my OTB results are indeed not accurate. I have trained the model again and obtains slightly better results on OTB-2015 (about 70.7% on OTB). You can test the new results as well as the new model.
I also check other benchmark results. The results on large-scale datasets (LaSOT, GOT, TrackingNet) are accurate.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants