FeatherNets_Face-Anti-spoofing-Attack-Detection-Challenge-CVPR2019

FeatherNets_Face-Anti-spoofing-Attack-Detection-Challenge-CVPR2019 copied to clipboard

Is the test data successful? The model training shows good results, but the single picture test is bad. Can someone help me?

9

Thank you very much.

S130111

Did you train FeatherNetB_bs32-ir with only IR image?

2

If no, how to train with only IR image?

mactiendinh

If isinstance(topk, int) is False, then in return you will catch : UnboundLocalError: local variable 'return_single' referenced before assignment

Katerina5649

请问，CASIA-SURF的数据集什么时候可以开放出来？

2

作者你好，我看很多作者和我一样，都卡在数据集上了，请问什么时候您能发开您论文中提到的数据集和标签数据？

HulkMaker

module ‘roc’ has no attribute 'cal_metric'

1

main.py 的validate函数出现这个问题，我分别试过roc-0.3.2和roc-0.2.3都会出现这个问题，求怎么解决？

aloha000

the Fisher150 Net could not be tared to ckpt.data

ezone1987

Private Dataset : where and how to get it?

3

I read "Our Private Dataset(Available Soon)": where and how to get it? Marco

marcoippolito

从百度网盘下载的预训练模型解压需要密码

1

你好，我从百度网盘下载了你们提供的预训练模型，但是解压的时候需要密码，请问这个密码是多少呢? 

zhouhao-learning

focal loss是否只使用了1024维的输出向量的前两个元素

3

您在论文里说直接输出1024维的特征向量，不使用全连接层，这样在计算focal loss时是不是只使用了向量的前两个元素。这样的话是不是会丢失很多信息？

LuckyDoggy

Which model is the best one

6

Thank a lot for your work. I tried to test your pretrained models, but I have been confused, because there are too many pretrained ones. According to your results on...

khanhnt