Skip to content
Models from paper Kišš, Martin, Michal Hradiš, and Oldřich Kodym. “Brno Mobile OCR Dataset.” International Conference on Document Analysis and Recognition, 2019.
Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.idea
conv_net
lstm_net
.gitignore
LICENSE
README.md
transcribe.py

README.md

B-MOD

Models from paper M. Kišš, M. Hradiš, and O. Kodym, “Brno Mobile OCR Dataset” in 2019 15th IAPR International Conference on Document Analysis and Recognition (ICDAR), IEEE, 2019. (arxiv)

We also provide a python script for transcribing text lines with our proposed networks. The script runs with parameters specifying directory with image to transcribe and also used network. For example

python3 transcribe.py --input-dir=/tmp/my_dir/ --net=LSTM

transcribes all .jpg and .png files in /tmp/my_dir/ using neural network with recurrent layers (network stored in lstm_net/). If you want to transcribe lines using netowrk without recurrent layers, use --net=CONV parameter instead.

You can’t perform that action at this time.