Introduction

This fork contains modifications to the hisdb runner class. These modifications were made in order to be able to work with our own dataset for the task of printed and handwritten text identification in mixed documents. To reproduce the U-net experiment described in the paper, install DeepDiva using the intructions further bellow and execute following command:

source activate deepdiva

python template/RunMe.py --runner-class semantic_segmentation_hisdb --output-folder ../output_yolo_try2 --dataset-folder datasets/printed-hw-seg/ --ignoregit --model-name unet --epochs 20 --experiment-name unet_ultra_final  --batch-size 16 --crop-size 256 --imgs-in-memory 3 --crops-per-image 500 --optimizer-name Adam --no-val-conf-matrix -j 16 --seed 1451368622

This will train, validate and test the model. Note that even though we tell DeepDIVA to train for 20 epochs, the model at epoch 6 will be chosen to run the tests, as further training overfitts the network.

DeepDIVA: A Highly-Functional Python Framework for Reproducible Experiments

DeepDIVA is an infrastructure designed to enable quick and intuitive setup of reproducible experiments with a large range of useful analysis functionality. Reproducing scientific results can be a frustrating experience, not only in document image analysis but in machine learning in general. Using DeepDIVA a researcher can either reproduce a given experiment with a very limited amount of information or share their own experiments with others. Moreover, the framework offers a large range of functions, such as boilerplate code, keeping track of experiments, hyper-parameter optimization, and visualization of data and results. DeepDIVA is implemented in Python and uses the deep learning framework PyTorch. It is completely open source and accessible as Web Service through DIVAServices.

Additional resources

Getting started

In order to get the framework up and running it is only necessary to clone the latest version of the repository:

git clone https://github.com/DIVA-DIA/DeepDIVA.git

Run the script:

bash setup_environment.sh

Reload your environment variables from .bashrc with: source ~/.bashrc

Verifying Everything Works

To verify the correctness of the procecdure you can run a small experiment. Activate the DeepDIVA python environment:

source activate deepdiva

Download the MNIST dataset:

python util/data/get_a_dataset.py mnist --output-folder toy_dataset

Train a simple Convolutional Neural Network on the MNIST dataset using the command:

python template/RunMe.py --output-folder log --dataset-folder toy_dataset/MNIST --lr 0.1 --ignoregit --no-cuda

Citing us

If you use our software, please cite our paper as (will be updated soon):

@inproceedings{albertipondenkandath2018deepdiva,
    archivePrefix = {arXiv},
    arxivId = {1805.00329},
    eprint = {1805.00329},
    author = {Alberti, Michele and Pondenkandath, Vinaychandran and Würsch, Marcel and Ingold, Rolf and Liwicki, Marcus},
    title = {{DeepDIVA: A Highly-Functional Python Framework for Reproducible Experiments}},
    year = {2018},
    month = {apr},
}

License

Our work is on GNU Lesser General Public License v3.0