crnn icon indicating copy to clipboard operation
crnn copied to clipboard

Convolutional Recurrent Neural Network (CRNN) for image-based sequence recognition.

Results 91 crnn issues
Sort by recently updated
recently updated
newest added

@bgshih I am trying to train crnn on a synthetic dataset that I generated. Along with english alphabets and digits (0-9) it contains special characters like $,@,# etc. I was...

For example the good result with short width ![image](https://cloud.githubusercontent.com/assets/16144935/22959941/1ca7138c-f375-11e6-8eee-0a8e56c12340.png) ![image](https://cloud.githubusercontent.com/assets/16144935/22959946/299b685e-f375-11e6-8114-76d407777d87.png) the bad result with long width ![image](https://cloud.githubusercontent.com/assets/16144935/22960009/a60b6c04-f375-11e6-8f63-33755cb436d7.png) ![image](https://cloud.githubusercontent.com/assets/16144935/22960017/ae42702a-f375-11e6-8519-eea0c0a1ca0a.png)

MDB_CORRUPTED: Located page was wrong type

我现在部署的pytorch warp-ctc 都是基于cpu的,我还能训练吗

My identification result is upside down, what reason is excuse me. Thank you @bgshih ![image](https://user-images.githubusercontent.com/39403315/50582211-a3108780-0e9b-11e9-8a76-2fa4d6a35eaa.png)

The link given in README.md is dead now. This PR updates the link of "CTPN+CRNN implementation" to a new one.

Hello , I have been searching everywhere to know how to create the label_list file ? lets say my Image_list file looks like image1.jpg image2.jpg image3.jpg so is it the...

hello,when i compile the fbthrift ,i meet some problems, can anybody give me some suggestion? thank you very much. i have already compile the folly success. the error message: /home/hey/torch/fblualib/folly/_build/include/folly/functional/ApplyTuple.h:45:40:...

I train the model in the dataset'Chinese text in the wild'(清华腾讯交通标志数据集),and test in the my own data(some text on traffic sign),the accuracy is 0,do you know the reason?And how to...

function loadAndResizeImage(imagePath) local img = Image.load(imagePath, 3, 'byte') img = Image.rgb2y(img) img = Image.scale(img, 100, 32)[1] return img end @bgshih