lyl8213/Plate_Recognition-LPRnet

为什么训练到30epoch左右就出现train_cost为0,而且val_cost没有减少?而且使用model8.24best里的模型进行测试时都是出现长度不一致,得不到测试结果?

YDS-FOREVER opened this issue · 8 comments

训练的现象如下:
Epoch 26/300, steps = 3822, train_cost = 0.000, logits_size = (24, 50, 67), val_cost = 12.472, val_ler = 0.146, time = 1.824s, learning_rate = 0.0009000000427477062 Epoch....... 26 Epoch 27/300, steps = 3969, train_cost = 0.000, logits_size = (24, 50, 67), val_cost = 13.910, val_ler = 0.148, time = 1.781s, learning_rate = 0.0009000000427477062 Epoch....... 27 Epoch 28/300, steps = 4116, train_cost = 0.000, logits_size = (24, 50, 67), val_cost = 13.581, val_ler = 0.156, time = 1.778s, learning_rate = 0.0008099999977275729 Epoch....... 28 Epoch 29/300, steps = 4263, train_cost = 0.000, logits_size = (24, 50, 67), val_cost = 13.747, val_ler = 0.156, time = 1.954s, learning_rate = 0.0008099999977275729

测试的现象如下:
len(original_list) 50 len(detected_list) 1 test and detect length desn't match time:0.2623302936553955 len(original_list) 50 len(detected_list) 1 test and detect length desn't match time:0.27623462677001953 len(original_list) 50 len(detected_list) 1 test and detect length desn't match time:0.2702765464782715 len(original_list) 50 len(detected_list) 1 test and detect length desn't match time:0.27227091789245605 len(original_list) 50 len(detected_list) 1 test and detect length desn't match

出现同样的问题,你解决了吗?

@yangdashi88 @zhaorui0907 @lyl8213 测试的时候,同样出现这样的问题,不知道该怎么解决的?
image

你用的什么样数据集,过拟合严重啊

@lyl8213 我用的就是你训练好的模型,我看前面几个提问的人都存在这个问题!不知道是哪里出现了问题呢?

哦,我放在train里面的图片数据是原始数据,你们自己需要进行augmentation(进行角度旋转,加噪声,缩放等等,可参考),我还生成了部分假车牌以补齐数据集中省份较少的车牌(用的这个里面的程序szad670401 / end-to-end-for-chinese-plate-recognition),当时是图片太多传的慢所以就没传augmentation之后的数据集,忘了说明

测试出现的现象我再看看

@lyl8213 好的,非常感谢你的解答

image
这边dd= session.run(decoded[0], test_feed)这句话run出来之后dd的第一个维度是0,所以出现不匹配的问题,不知道是decoded出问题了没?

klq-Y commented

请问这个问题在代码里面咋改啊?还是要改数据集?