simple-railway-captcha-solver

simple-railway-captcha-solver copied to clipboard

可不可以用簡體字啊，我大多數看不懂的。 （可不可以用简体字，我大多数看不懂的。）

zhengyo

可不可以用簡體字啊，我大多數看不懂的。 （可不可以用简体字，我大多数看不懂的。）

zhengyo

help me how to read this 

ott18

Update README.md

1

Change in readme.mc

Ashishpurbey

in training and valid code, if you specify metrics=["accuracy"] in the model.compile(), then the history object will have the keys as 'accuracy' and 'val_accuracy'. While if you specify it as...

coseto6125

Can you please explain the issue you are trying to address through this?. because I am not getting what you are trying to solve through this. Thank you

Mojahid-Ahmad

>out = Conv2D(filters=32, kernel_size=(3, 3), padding='same', activation='relu')(out) >out = Conv2D(filters=32, kernel_size=(3, 3), activation='relu')(out) 請問為甚麼這裡每次都先做兩次Conv再做一次Pooling? 看過的教學都是Conv一次然後做一次Pooling 找不到相關技巧的說明

AoiYamada

也許可以做個CRNN以適應不定長度的驗證碼 另外再提供個 GAN 範例就更好了。 如果有GAN有問題的話，可以收PR而修正。

mason276752

您好，我看到在`train_cnn_imitate_6.py`里，关于`train_label`的代码，最后得到的`train_label`就是一个长度6的数组 ``` traincsv = open('./data/6_imitate_train_set/captcha_train.csv', 'r', encoding = 'utf8') read_label = [toonehot(row[1]) for row in csv.reader(traincsv)] train_label = [[] for _ in range(6)] for arr in read_label: for index in...

die9o