method: 4Paradigm-Data-Intelligence2019-06-03

Authors: Feng Cheng, Lixin Gu, Qingjie Liu, Feng Han, Jingtao Han

Description: The detection model and recognition model are trained separately.
Detection model: Based on Mask-RCNN. multi-scale. Train-set: 2017 MLT task1 train-set.
Recognition model: Based on Transformer with backbone ResNet50. A voting process is done to identify the language of recognized transcript. Train-set: 2017 MLT task2 train-set & 2019 MLT task2 train-set & 2019 MLT Synthetic dataset.

method: CLOVA-AI2019-06-04

Authors: Bado Lee, Youngmin Baek, Hwalsuk Lee

Description: Additional head on Character-level text detection with model distillation. A pretrained detector is used.

CLOVA-AI team, Naver Corp.

method: Ashwaq2019-05-10

Authors: Ashwaq

Description: FCN

Ranking Table

Description Paper Source Code
DateMethodHmeanPrecisionRecallAverage Precision
2019-06-034Paradigm-Data-Intelligence75.23%79.26%71.60%56.65%
2019-06-04CLOVA-AI68.31%74.52%63.06%54.56%
2019-05-10Ashwaq58.11%62.44%54.34%40.61%
2017-06-30SCUT-DLVClab258.08%71.78%48.77%41.42%
2017-06-30TH-DL39.37%58.58%29.65%24.54%

Ranking Graphic

Ranking Graphic