Closed Gary-code closed 1 year ago
Congratulations on the paper accepted by CVPR. Here is a question, the iod features can not concatenate with the bottom-up features, because the object num seems not equal? After carefully reading the paper, I think each object detected by faster-RCNN has an unbiased feature. Looking forward to your reply.
您好,请问您iod 特征在哪里下载的啊,是作者给的那个失效的百度云链接吗? :)
Congratulations on the paper accepted by CVPR. Here is a question, the iod features can not concatenate with the bottom-up features, because the object num seems not equal? After carefully reading the paper, I think each object detected by faster-RCNN has an unbiased feature. Looking forward to your reply.
Hello, do you have the cloud disk file shared by the author?
Congratulations on the paper accepted by CVPR. Here is a question, the iod features can not concatenate with the bottom-up features, because the object num seems not equal? After carefully reading the paper, I think each object detected by faster-RCNN has an unbiased feature. Looking forward to your reply.