Ian Ren
Hi, I also encountered this bug, it seems some details in "DCN" lib are not handled well enough to avoid this problem, so when it comes to the last step...
@techzhou Hi, how about the accuracy after u implement dropout layer? did it increase a little bit? I just think dropout may help. thanks!
@wangkangnian I used CUB-200-2011 dataset. I will post my part of no-random initialization, if it helps.
@wangkangnian @guyibang I have posted my part of non-random initialization. I wish it may help you guys a bit.😊
thanks a lot! I see it
@songdejia
hi, I find that the section 3.3 **layer initialization** in the paper is not implemented in your code, and I am a little confused about this section, maybe it`s not...
@pengshiqi Not yet... I am working on it these days... and how does the ratio of loss influence the result? have u ever tried others? thanks
@pengshiqi Also, using tencrop on the pics makes a difference, but only limited...
@songdejia hi， would you plz post your loss value when you get 85.14% in test set? thanks, Ian