anoopyear2020 / sroie2019-1

SROIE 2019

Installation

Clone this repo under ~/Documents:

git clone --recurse-submodules https://github.com/loveorchids/sroie2019 ~/Documents/sroie2019

Requirement

Python: 3.5.2 or higher

pip install -r requirements.txt

Prepare Data

1. Download dataset from SROIE 2019 to PATH_TO_DATASET
2. Create path ~/Pictures/dataset/ocr/ and move SROIE dataset under it

cd ~/Pictures
mkdir dataset
cd dataset
mkdir ocr
cd PATH_TO_DATASET
# rename dataset
mv PATH_TO_DATASET/DATASET_NAME PATH_TO_DATASET/SROIE2019
# move dataset
mv -r PATH_TO_DATASET/SROIE2019 ~/Pictures/dataset/ocr/

Training

# Run in Terminal
cd ~/Documents/sroie2019/researches/ocr/textbox
python3 textbox.py

Testing

# Run in Terminal
cd ~/Documents/sroie2019/researches/ocr/textbox
python3 tb_test.py (some args)

Pretrained Model for Task1

The F1-score on SROIE 2019 test set is expected to be 94%.

https://drive.google.com/open?id=1uZnBuLm_DLKwNukYhwbrbDxUmLa2XtPe