sirius-ai/LPRNet_Pytorch

Issues

• 报错：ValueError: num_samples should be a positive integer value, but got num_samples=0

  #45 opened 4 years ago by Xccelrator
  5

• ZeroDivisionError: float division by zero

  #62 opened 3 years ago by VSERunKNOWN
  2

• 你们测试的正确率都有多少

  #44 opened 4 years ago by XieHui1995
  31

• 12 types of license plate recognition, implemented by C++ TneosrRT, with an accuracy rate of up to 99.5%!!!

  #92 opened 4 months ago by pcb9382
  0

• test_accuracy 0.0

  #75 opened 2 years ago by kankanjiuzou123
  1

• 我在训练集放了3000张，测试集里面放了200张，但是我无论怎样放总会出现[Info] Test Accuracy: 0.0 [0:120:0:120]，也就是tp和tp_2总是为0，我想问一下这是为什么

  #91 opened 4 months ago by nanlidashi
  0

• Is it possible to compute the accuracy output from the LPRNet model?

  #85 opened 10 months ago by tscu-vid
  1

• Enhancement: Expanding LPRNet's Recognition Capabilities to Global License Plate Varieties

  #90 opened 5 months ago by yihong1120
  1

• 请问一下数据集怎么标注

  #88 opened 7 months ago by programcj
  1

• 我的loss从2、3降到0.5左右，acc从0加到0.2，然后loss开始变成负数了，很奇怪！acc也从0.2突变成0

  #86 opened 10 months ago by zhaojiankuan
  1

• 12 types of license plate recognition, implemented by C++ TneosrRT, with an accuracy rate of up to 99.5%!!!

  #89 opened 6 months ago by pcb9382
  0

• train loss gets extraordinary higher sometimes

  #87 opened 7 months ago by jeungmin717
  0

• 中途为啥test accurary为0.0啊？

  #74 opened 2 years ago by Vergissmeinnic
  5

• 数据集

  #84 opened a year ago by daijie1223
  1

• 提供50w中文车牌数据集

  #70 opened 2 years ago by 123Vincent2018
  1

• Questions about format of licenseplate

  #83 opened a year ago by jeungmin717
  0

• lose from 0.0001-->5.x

  #61 opened 3 years ago by wavelet2008
  1

• license plate recognition

  #81 opened a year ago by ZhaoMonica
  0

• 省份识别精度不高

  #57 opened 3 years ago by duorouputao
  6

• maxpool3d 在tensorrt中不支持

  #66 opened 2 years ago by leaf918
  2

• model.train()和model.eval()

  #80 opened a year ago by petervxy
  0

• Errors in loading state_dict (size mismatch)

  #71 opened 2 years ago by dubdu
  3

• ncnn推理，已经C++结合ncnn实现lprnet模型推理，有需要可以交流

  #78 opened 2 years ago by Ayers-github
  7

• Acc = Tp * 1.0 / (Tp + Tn_1 + Tn_2) ZeroDivisionError: float division by zero

  #65 opened 3 years ago by SUTwu
  8

• 您好，我在测试的时候出现AttributeError: 'NoneType' object has no attribute 'shape'的问题

  #64 opened 3 years ago by SUTwu
  10

• 图片名字有中文字的时候会报错no module named shape，可是改了名字之后不就没标签了吗？请问该怎么办

  #77 opened 2 years ago by littleloverlalala
  2

• 图片尺寸、字符类别、最大字符数都需要改，应该怎么用？

  #79 opened 2 years ago by nistarlwc
  1

• train的那个文件中没有labels的路径吗？所以不用labels训练？

  #63 opened 3 years ago by frandoLin
  5

• 这个模型可以识别双层车牌么？

  #60 opened 3 years ago by hezinan
  9

• Accuracy 0.0, when i training custom dataset.

  #76 opened 2 years ago by Lightmourne
  4

• 车牌识别商用级别数据集

  #73 opened 2 years ago by daijie1223
  2

• 数据集命名

  #72 opened 2 years ago by Smiler-X
  0

• train_LPRNet.py文件中的Greedy_Decode_Eval函数，使用测试集评估模型的一些问题

  #59 opened 2 years ago by regainOWO
  0

• > 用预训练的模型测试所给的测试集，怎么只有0.9的正确率，再训练了自己的数据集之后就只有0.74了。。。

  #50 opened 3 years ago by hanrui15765510320
  4

• 请问lprnet的输入如何改成Dims3呢

  #69 opened 2 years ago by OmiLiuyueyue
  0

• Reduced vs basic model

  #49 opened 3 years ago by barrypitman
  3

• 这个训练的权重是使用蓝牌数据集训练的吗？

  #68 opened 2 years ago by Tomhardy13
  0

• 训练次数如何修改？

  #54 opened 3 years ago by luxianlebron
  1

• 请问车牌号码字符的置信度分数如何获得呢？

  #67 opened 2 years ago by Sunsamar
  0

• Hi, 感谢你的贡献, 有个小问题, 怎么检测手机拍的照片呢

  #55 opened 3 years ago by zhoudelong
  1

• 可提供车牌黄绿蓝数据集

  #58 opened 3 years ago by daijie1223
  0

• 可提供数据支持

  #51 opened 3 years ago by daijie1223
  0

• Error in train_LPRNet.py

  #56 opened 3 years ago by saumyapatel17
  1

• how to feed the labels/annotation of the images during training?

  #53 opened 3 years ago by tsan1
  1

• Does input_size value (94x24) can change to other value?

  #52 opened 3 years ago by Tienson0907
  0

• datasets

  #42 opened 4 years ago by ccsone
  1

• 这是我自己加了一些字符后训练后测试的结果，简直是惨不忍睹呀！

  #48 opened 3 years ago by rawli
  2

• Normalisation

  #47 opened 3 years ago by NgoJunHaoJason
  0

• Hyperparameters

  #46 opened 3 years ago by NgoJunHaoJason
  0

• 训练过程中，forward输出全部都是nan

  #43 opened 4 years ago by onnx20
  1