search

researchmm / TracKit

[ECCV'20] Ocean: Object-aware Anchor-Free Tracking
MIT License
613 stars 97 forks source link

TrackingNet support #61

Closed SilvioGiancola closed 3 years ago

SilvioGiancola commented 3 years ago

Dear team,

Thank you for sharing your code, it is great for the tracking community. Do you have any plan on including experiments/benchmarks on TrackingNet?

Please let me know,

Best,

penghouwen commented 3 years ago

Hi @SilvioGiancola ,

Glad to hear from you. We're interested in integrating Ocean and TracKit toolkit into the wide-used TrackingNet. @JudasDie Zhipeng, would you help to discuss with @SilvioGiancola on the integration? Thanks.

Thanks, Houwen

JudasDie commented 3 years ago

Dear team,

Thank you for sharing your code, it is great for the tracking community. Do you have any plan on including experiments/benchmarks on TrackingNet?

Please let me know,

Best,

Hi, thanks for your interest. The results and testing script would be updated this week.

SilvioGiancola commented 3 years ago

Thanks @penghouwen and @JudasDie for your feedback! Any update on your side?

If that can be of any help, a pip package for TrackingNet is now available. It integrates functions to retrieve the IDs of specific tracking sequences in training and testing as well as API calls to download the sequences.

If you are facing any issue in using those API, please feel free to reach out.

Best,

JudasDie commented 3 years ago

Thanks @penghouwen and @JudasDie for your feedback! Any update on your side?

If that can be of any help, a pip package for TrackingNet is now available. It integrates functions to retrieve the IDs of specific tracking sequences in training and testing as well as API calls to download the sequences.

If you are facing any issue in using those API, please feel free to reach out.

Best,

I plan to integrate a testing script like lib/core/eval_otb.py. Is that ok for your need? (sorry for the late update, I'm busy with other projects recently)