online_khatt

The train script and feature preparation were taken from DeepOnKHATT project. The goal is to improve/optimize the code and fine-tune the model on other dataset

Environment Setup

conda env create -f environment.yaml

Data Preparation

All handwriting samples have to be in the following format

679.785826771654 70.0346456692913 0
679.785826771654 70.0346456692913 0
679.181102362205 68.4850393700787 0
678.727559055118 67.8047244094488 0
678.047244094488 67.7291338582677 0
................. ............... .
................. ............... .
................. ............... .
................. ............... .
676.573228346457 70.2236220472441 1

The first column is x coordinate, the second column is y coordinate and the third column is pen up indicator. \

All samples files should have label file (examples and jupyter notebook provided in features directory)

Configuration

General configuration can be found in neural_network.ini file

Training

To start training from scratch run the following command: \

python src/models/train.py --config neural_network.ini

make sure that neural_network.ini is under src/configs/ directory

Fine-tuning

To load the pre-trained model and continue training run the following command: \

python src/models/train.py --config neural_network.ini --path models/model.ckpt-i