redlt

Questions about reimplementing the experiments in your paper.

3

Thanks for your great contribution. I have read your paper, but have some questions about the experiments in Table 1. 1. The replacement of the predicted results is only implemented...

想请教一下attention2d里的temperature是干嘛的？

2

一直限制attention的输出大小，这是warmup的手法吗？为什么是加在attention后面而不是卷积后面呢？

attention map as shown in your paper Figure 6. attention distance as Figure 7(right).

why here it has frozen bn twice?

2

in FCOS.Pytorch/model/fcos.py ``` def freeze_bn(module): if isinstance(module,nn.BatchNorm2d): module.eval() classname = module.__class__.__name__ if classname.find('BatchNorm') != -1: for p in module.parameters(): p.requires_grad=False ``` Since module.eval() has frozen bn, why you additionally set...

FOCS has been accepted by maskrcnn_benchmark, but I dont know how to use stitcher in FCOS. Can I just copy FCOS's config to this project?

这里应该是激活函数的导数吧？激活函数的符号前文中是\sigma；感叹号变为1

Could anyone explain why don't use validation for baseline?

1

If we don't use validation and don't calculate accuracy during training, how can we know when to stop and how it performs in the training set, and whether it has...

Could you please give me some suggestions about building such environment?

5

I downloaded your code, but I don`t know how to run this demo for I am unfamiliar with cpp. I have installed vs2015, qt 5.8.0,cuda, opencv, and downloaded the tensorflow...