I tried to do it and found that the result I obtained was incorrect (mPrec/mRec/mCov/mWCov are different from the SoftGroup paper, but ap/ap50/mIou/Acc are same), especially in pred_ Instance and gt_instance I am very confused about the matching of instances. I hope you can provide some code in this area or provide some guidance. I would greatly appreciate it!Thank you very much!
it should be a error in pred_instance for matching gt_instance in the code wrote by myself,so i hope you could share the code for evaluating mPrec mRec mCov mWCov,thank you in advance!
I noticed in issue # 46 that you mentioned using https://github.com/WXinlong/ASIS/blob/master/models/ASIS/eval_iou_accuracy.py to evaluate the metrics like mPrec mRec mCov mWCov
I tried to do it and found that the result I obtained was incorrect (mPrec/mRec/mCov/mWCov are different from the SoftGroup paper, but ap/ap50/mIou/Acc are same), especially in pred_ Instance and gt_instance I am very confused about the matching of instances. I hope you can provide some code in this area or provide some guidance. I would greatly appreciate it!Thank you very much!![Screenshot_2023-09-07-11-15-38-896_com android browser](https://github.com/thangvubk/SoftGroup/assets/97398621/2832d5b2-0ac4-4008-a952-db8144b57404)