happynear
verification problem
您好:
我是Caffe的使用者，我使用CASIA database 訓練出identification 達到82%，但是我想把特徵擷取下來做JB，但是一直無法把辨識率做到90%已上，Feature extract 我是用caffe提供的matlab的函式去做截取的，我在想會不會是擷取特徵上的問題，請問大師你是用哪個方法擷取特徵的呢?
您好： 我最近也在实现DeepID2的verification，希望可以做一个实际应用。但是在原文中提到需要使用25个CNN提特征，然后再进行PCA，我觉得这样在实际应用中太消耗时间和计算资源。所以想请问你一下，你在实现你的matlab版本的时候是怎么做的，希望你可以提供一些帮助。谢谢~
Thank!
But i don't understand why you set 0.0078125 scale in CASIA_train_test.prototxt. Why is not 0.00390625.Should it normalized -1 to 1 not 0 to 1? But input image is 0 - 255 , if scale set 0.0078125 it get 0-2 isn't it? So why scale is 0.0078125. Do you do any image process before you train model? Thank your answer!
@m10303430 , I have also minused the mean image. So every pixel's range is [-1 1].
Thank
But i accuracy still can't achieve as yours , my database is CASIA normalized version and i take your CASIA_train_test.prototxt and mnist_siamese_solver.prototxt to train CNN.But relu layer type i use "ReLU" not your "Insanity".Validation set accuracy rate is about 72%. But when i test verification i only get 79%. The problem is relu layer? Or you can get me some suggestion.