xuannianz / keras-ctpn

CTPN (Connectionist Text Proposal Network) implementation in Keras

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

keras-ctpn

This is an implementation of CTPN on keras and Tensorflow. The project is based on matterport/Mask_RCNN and eragonruan/text-detection-ctpn. Thanks for their hard work.

Test

  1. I trained by concatenating two different datasets (icdar 2019 art and rctw 2017). There are 8964 images for training and 2232 images for validation.
  2. The best evaluation result is
recall precision hmean
0.6886 0.8677 0.7678
  1. Pretrained model is here. baidu netdisk extract code: ezcj
  2. python3 inference.py to test your image.

Train

build dataset

  • First you need to split the annotations into small bboxes whose width is 8px and save them into a txt file for each image.
  • python3 split.py to split annotations in icdar2019 art dataset.
  • python3 split_rctw.py to split annotations in rctw 2017 dataset.
  • python3 build_dataset.py to save the images and annotations into hdf5 files so that the model can be trained very fast.

train

  • python3 train.py to start training.

Evaluate

  • python3 eval.py to evaluate.

Improve

  • data augmentation
  • deep backbone network
  • fpn

About

CTPN (Connectionist Text Proposal Network) implementation in Keras


Languages

Language:Python 100.0%