Note: If you use this code, please cite:

Deng, Dan, Haifeng Liu, Xuelong Li, and Deng Cai. "Pixellink: Detecting scene text via instance segmentation." In Thirty-Second AAAI Conference on Artificial Intelligence. 2018.

Liu, Yang, Zhaowen Wang, Hailin Jin, and Ian Wassell. "Synthetically supervised feature learning for scene text recognition." In Proceedings of the European Conference on Computer Vision (ECCV), pp. 435-451. 2018.

-- Prerequistes:

• Python 3.6
• Tensorflow-gpu >= 1.1
• opencv2
• setproctitle
• matplotlib
• pylib: https://github.com/dengdan/pylib.git

--Download pretrained model:

Download : https://drive.google.com/open?id=1PuLCYVG457UOFzWHz4GuerTzWABZR0b6

DETECTION: unzip pixel_link_vgg_4s.zip into ${pixel_link_root}/model/

RECOGNITION: put 0_480000.pth into $attention_net_root$/

-- Usage:

DETECTION:

1. Add the path of $pixel_link_root$/pylib/src to your PYTHONPATH

   export PYTHONPATH=${pixel_link_root}/pylib/src:$PYTHONPATH

2. Test on Images in demo.txt:

./scripts/test_any.sh 0 model/conv3_3/model.ckpt-38055 $pixel_link_root$ /demo.txt

3. It will generate detection result called Detection.pkl, which can be used as the input to the recognition model.

RECOGNITION:

1. Recognize the text from the detected boxes:

python Recognition_yang.py --detection_path $pixel_link_root$/Detection.pkl --gpus 0

2. The results will be saved in the file: $attention_net_root$/Recognition.pkl