Open jerichosu opened 2 years ago
Based on ReadMe:
"Train on CrowdHuman, MOT17, Cityperson and ETHZ, evaluate on MOT17 train."
So are you guys evaluating results on training dataset? Or this is just a typo? Same issue on other pre-trained models introduction.
Based on ReadMe:
"Train on CrowdHuman, MOT17, Cityperson and ETHZ, evaluate on MOT17 train."
So are you guys evaluating results on training dataset? Or this is just a typo? Same issue on other pre-trained models introduction.