Recall@1% on Oxford is higher than that in the paper.

Open banxiyan opened this issue 3 years ago • 1 comments

Thanks for your great work, the recall@1% on Oxford is higher than that in the paper, which is trained by the default config in your repository. Is it normal?

Sep 30 '22 00:09 banxiyan

Thanks for your great work, the recall@1% on Oxford is higher than that in the paper, which is trained by the default config in your repository. Is it normal?

hi, I have the same issue, I trained evaluated the network by default and got 87.39% recall, do you have any idea about that? By the way, i find in the file "train_pointnetvlad", the global variable "HARD_NEGATIVES={}" is never updated which means it is useless, is there something wrong here?

Oct 07 '22 08:10 Master-cai