naver / r2d2

Other
461 stars 86 forks source link

Some details about Oxford dataset #39

Closed ThompsonHe closed 3 years ago

ThompsonHe commented 3 years ago

Hi, sorry for bothering, when I try to reproduce the network's performance on oxford dataset, I was stuck by the process of using DoG detector to detect keypoints on an image. I am wondering whether the DoG detector is implemented by yourself or you just resort to some open source code(If so, could you please provide me the link?). Thank You!!!

jerome-revaud commented 3 years ago

You can find many implementations of the DoG detector, e.g. in OpenCV or VLFeat.