YujiaoShi / IBL

14 stars 2 forks source link

Fine-grained cross-view codes missing (maybe)? #5

Open liguopeng0923 opened 1 year ago

liguopeng0923 commented 1 year ago

Hi yujiao,

I cannot find the Fine-grained cross-view codes (the second stage as mentioned in Sec.4 of your paper), so I cannot understand how it work in training or testing time.

If I ignore corresponding codes, please note that for me. If not, I will sincerely appreciate if you can provide it.

Best wishes.

liguopeng0923 commented 1 year ago

Besides, It's hard to understand how to evaluate the three groups in Fine-grained cross-view Image matching (Sec 5.3). For examples, what are the inputs for them? I guess the experiments in this section are based on random images selected by you and evaluated by your companions (user study)?