Angelina braille reader

A modification of IlyaOvodov/AngelinaReader

Requirements

• Git with Git LFS
• Python 3.6-3.8

Setup for development

Installation

1. Run from the command line (MacOS / Linux / CygWin / Msys2):

   git clone --recursive https://github.com/braille-systems/AngelinaReader.git
   cd AngelinaReader
   git clone --recursive https://github.com/braille-systems/brl_ocr.git
   python -m pip install --upgrade pip
   python -m pip install virtualenv
   python -m venv env
   

   Substitute python with your system's Python3 command, which may be python, py or python3

2. Activate a virtual environment:

   source env/bin/activate

   (on Linux/MacOs) or

   .\env\Scripts\activate

   (on Windows)

   Note: to work on this project, you need to re-activate the environment every time you restart the machine or open a new terminal session. To verify that you've successfully activated the environment, run which python (or where python on Windows). You should see <...>\AngelinaReader\env\Scripts\python.exe on Windows (and something alike on other platforms) at the top of the list.

3. Install the dependencies:

   python -m pip install --upgrade pip
   pip install torch===1.5.0 torchvision===0.6.0 -f https://download.pytorch.org/whl/torch_stable.html
   pip install -r requirements.txt
   pip install -r model/requirements.txt
   

Running pre-trained inference

To ensure that you've done everything correctly, you may launch a pre-trained model:

wget -O weights/model.t7 http://angelina-reader.ovdv.ru/retina_chars_eced60.clr.008
python run_local.py brl_ocr\data\unlabeled\golubina\IMG_2503.JPG

Training

Execute the file: python model/train.py If you're getting CUDA errors, you may change device="cuda:0" to device="cpu" in model/params.py

This will run for a while. Results will appear under NN_results folder.

Labels

See brl_ocr/README.md#working-with-labels.

It's OK to install labelme in the same environment as other dependencies in this project.