/ndcg_wordspotting.pytorch

Primary LanguagePythonMIT LicenseMIT

Learning to Rank Words: Optimizing Ranking Metrics for Word Spotting

Introduction

Installation

git clone https://github.com/priba/ndcg_wordspotting.pytorch.git
cd ndcg_wordspotting.pytorch
conda env create -f environment.yml
conda activate ndcgws

Train

python main.py --data_path /path/to/GeorgeWashington/

Test

python main.py --data_path /path/to/GeorgeWashington/ -t -l
/path/to/trained/weights/

Data

George Washington

IIIT5K

Citation

@inproceedings{riba2021sndcg,
    author = {Riba, Pau and Molina, Adrià and Gomez, Lluis, and Ramos-Terrades,
    Oriol and Lladós, Josep},
    title = {Learning to Rank Words: Optimizing Ranking Metrics for Word Spotting},
    booktitle = {Proceedings of the International Conference on Document Analysis and Recognition (ICDAR)},
    year = {2021}
}