CRNN_Chinese_Characters_Rec

Characters Recognition

A Chinese characters recognition repository based on convolutional recurrent networks.

Performance

Recognize characters in pictures

Data

Synthetic Chinese String Dataset

1. Download the dataset in here
2. Put char_std_5990.txt in lib/dataset/txt/
3. Download the preprocessed labels in here (password:w877)
4. And put train.txt and test.txt in lib/dataset/txt/

Data

Synthetic Chinese String Dataset

1. Download the dataset in here
2. Put char_std_5990.txt in lib/dataset/txt/
3. Download the preprocessed labels in here (password:w877)
4. And put train.txt and test.txt in lib/dataset/txt/

Dev Environments

1. WIN 10 or Ubuntu 16.04
2. pytorch 1.4.0 (may fix ctcloss)
3. yaml
4. easydict

Train

   [run] python train.py --cfg lib/config/360CC_config.yaml

loss curve

   [run] cd output/360CC/crnn/xxxx-xx-xx-xx-xx/
   [run] tensorboard --log_dir log

loss overview

Demo

   [run] python demo.py --image_path images/test.png --checkpoints output/checkpoints/mixed_second_finetune_acc_97P7.pth

References

• https://github.com/meijieru/crnn.pytorch
• https://github.com/HRNet