ilovin lstm_ctc_ocr issues

ilovin / lstm_ctc_ocr

Use CTC + tensorflow to OCR

https://ilovin.github.io/2017-04-06/tensorflow-lstm-ctc-ocr/

354 stars 140 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

time_steps.append(nw//cfg.POOL_SCALE+cfg.OFFSET_TIME_STEP)

#56 SolinTao closed 5 years ago
1
test.sh出错

#55 yangjx54 opened 6 years ago
0
BN

#54 longmao-yiran opened 6 years ago
1
./test.sh error

#53 boris-lb opened 6 years ago
7
关于图像变长字符识别的问题

#52 longmao-yiran closed 6 years ago
3
字符的长度应该是63个吧

#51 fancyerii closed 6 years ago
2
请问训练数据字符个数长度不一致的时候有没有什么好的训练方式？

#50 billyzju closed 6 years ago
5
why num classes is __C.NCLASSES = len(__C.CHARSET)+2

#49 seeyourcell closed 6 years ago
2
ctc和tensorflow绑定以后运行出了问题， undefined symbol: _ZNK10tensorflow14TensorShapeRep11DebugStringB5cxx11Ev

#48 zhangtao22 opened 6 years ago
2
ctc

#47 zhangtao22 closed 6 years ago
0
运行代码时崩溃：段错误（核心已转储）

#46 Shirhe-Lyh closed 6 years ago
0
i can't make the warpCTC tensorflow_binding environment

#45 jiachen0212 closed 6 years ago
0
why there comes out negative loss?

#44 shazhumao closed 6 years ago
2
IOError: unknown file format

#43 gccrpm closed 6 years ago
0
undefined symbol: _ZTIN10tensorflow8OpKernelE

#42 gccrpm closed 6 years ago
0
ImportError: No module named utils.data_util

#41 gccrpm closed 6 years ago
3
一个小问题

#40 phamhe closed 6 years ago
2
Is this a scene text recoginizer?

#39 arsalan993 closed 6 years ago
1
Loss 小于0......

#38 Mendel1 closed 6 years ago
0
got AttributeError when training

#37 ugtony closed 6 years ago
3
Using Tensorflow CTC on Dev and Beta instead of Warp CTC

#36 selcouthlyBlue closed 6 years ago
2
there are no space “ ” in the labels，why add a space class in num_classes？

#35 jacobunderlinebenseal closed 6 years ago
4
关于修改为tf.nn.ctc_loss后收敛变慢的问题

#34 phamhe closed 6 years ago
3
关于master版本的代码无法收敛的问题

#33 ilovin opened 6 years ago
4
master standard 跑100多轮还是acc=0

#32 BrookShuihuaLee closed 6 years ago
0
master standard 跑100多轮还是acc=0

#31 BrookShuihuaLee closed 6 years ago
2
related paper

#30 hanghang2333 opened 6 years ago
2
环境搭建出现问题，作者你的环境具体是哪些版本？

#29 hookover closed 6 years ago
7
accuracy = 0.000 after Epoch 10000/10000

#28 AllenDuane closed 7 years ago
5
What's the difference between the dev version and beta version of the code? I did not find it

#27 gnnbest closed 7 years ago
0
question about a parameter

#26 cassiaaaaaa closed 7 years ago
4
Master branch cannot converge

#25 GabrielLin closed 7 years ago
2
i used chinese to run the code,but there has something wrong? acc is o,avg_train_cost <0?

#24 dotsonliu closed 7 years ago
7
i use master version ,i see your acc is 96%，i use 12800 images, the acc is only 76%, how many images you used?

#23 dotsonliu closed 7 years ago
1
i have run the code，the rusult acc is 86%，my goal is 90%+ ,how to impove this?

#22 dotsonliu closed 7 years ago
1
can i use chinese to train the model ?, in english ,you use 10000+image to train the model, can the acc Achieve 99%？ if i want to Achieve 99%, can you give me some advice. thank you

#21 dotsonliu closed 7 years ago
1
@rolai can i use chinese to train the model ?

#20 dotsonliu closed 7 years ago
0
OSError: cannot open resource

#19 MyraBaba closed 7 years ago
6
warpCTC: when I run python lstm_ocr.py to training it just use CPU not use GPU

#18 gnnbest closed 6 years ago
2
wrapctc

#17 zwl22 closed 7 years ago
1
WrapCtc的GPU和CPU版本

#16 zwl22 closed 7 years ago
1
will this work with dual line text?

#15 gewenpulan closed 6 years ago
3
两个疑问

#14 BrianGong closed 7 years ago
2
why printed “decode” same？

#13 BrianGong closed 7 years ago
4
Update train.sh

#12 Naruto-Sasuke closed 7 years ago
0
makedirs makes manually creating `data` folder free

#11 Naruto-Sasuke closed 7 years ago
0
Accuracy 0 with latest code

#10 ashish1405 closed 7 years ago
15
How to deal with different size of images?

#9 Naruto-Sasuke closed 6 years ago
5
ValueError: zero-size array to reduction operation maximum which has no identity

#8 Naruto-Sasuke closed 7 years ago
3
Cannot run the code of the master branch

#7 Naruto-Sasuke closed 7 years ago
0